Does Black actually have more control than White over the direction of the opening??

The opening is a combination of both player's choices. The only real control white has is the first move and will need plans for any possible replies

But when you play the Sicilian you have to know the English, Bg5, Closed (fianchetto), Grand Prix, Morra Smith, Opocensky, c3 sicilian, Canal, and... Bowdler, not to mention h3/g4/Rg1 ideas that are far better in practice and by computer analysis than anyone cares to admit.

In a certain sense yes, if white just plays 1. e4 or 1. d4 there's more diversity to blacks first move responses... so it would seem white has to know more theory than black. But only because white usually plays in such a predictable manner. If white were to play any of the perfectly viable other moves... like 1. b3, 1. Nc3, 1. f4, 1. b4, 1. c3, 1. Nf3, 1. g3, 1. e3, 1. c4... well, then no.

Though you also have to remember that white still has the initiative, even within the various moves black plays. This doesn't matter as much in the sicilian - black eventually chooses the major variation in the sicilian... it's one of the biggest advantages of the sicilian. But in e4e5, for example, white usually chooses the main line - ruy, italian, vienna, scotch, etc.. And in most lines this is how it works. So that kind of offsets your observation.

The french may be the other e4 response where I feel black maintains more control over the game than white. Unless white is willing to cede equality. But in the caro-kann... and any other e4 response you can name, it's definitely white in control.

Players need to be more creative as white, but they just aren't, they get too caught up with pursuing the objectively ideal lines. And usually there will be plenty of opportunities for you to take control of the game regardless. But yeah as black you can really just focus on d4 / e4 and ignore everything else for a long time.

It's an interesting question. I don't think that either side exhibits greater control over the direction of the opening. Black may have the first opportunity to dictate the course of the game but white isn't lacking opportunities either.

Take the Sicilian Defense for example, outside of 2.Nf3, white has 5-6 other reasonable choices on the second move. In fact, white has so many options to avoid the main lines that 500+ page books have been written devoted to that subject matter.

White have 19 fine first moves.

Short answer is no.
White could choose to send the game in very different directions from move one depending on if s/he plays 1. e4 or, say, 1. b3.

If white plays 1. e4, then black can choose from many defenses with differing characters, but that's after white has already sent the game a certain direction with 1. e4.

No.

People are confused by that due to naming. It is equally an issue for both players. Let's look at a few examples, and this goes for defenses and attacks.

Let's start with your opening - The Sicilian.

Let's say, hypothetically, that you play the Najdorf against the Open Sicilian, and you are facing an e4 player.

HE must know:

ONE response to 1...e5

ONE response to 1...e6

ONE response to 1...c5

ONE response to 1...c6

Now Let's say that line against 1...c5 is the Open Sicilian. He needs ONE line against EACH Open Sicilian. So Let's say it is 6.Bg5 for the Najdorf.

YOU on the other hand need to know:

6.Bg5, 6.Bc4, 6.Be3, 6.Be2, 6.f4, 6.h3, 6.Rg1, 6.a4, 6.Bd3, and I am sure a few other minor lines.

Then, in addition to that, you need something against the Closed, Grand Prix, Moscow, Alapin, Delayed Alapin (1.e4 c5 2.Nf3 d6 3.c3 DOES NOT TRANSPOSE unlike other Sicilian move 2 responses), Morra Gambit, Wing Gambit, Delayed Wing Gambit, and other offbeat Sicilians like 2.a3, etc. White does not need to know all that. He needs to know pieces of other openings.

The same goes for attacks. I play the Trompowsky Attack. That sounds like a game White controls. Not at all.

As White, I need to know the following:

The Trompowsky Attack (1.d4 Nf6 2.Bg5) with 2...Ne4, 2...c5, 2...d5, 2...e6, 2...g6, and some offbeat lines.

The Levitsky Attack (1.d4 d5 2.Bg5) or else something else like the London, Colle, or Queen's Gambit

Something against 1...e6, 1...g6, 1...d6, 1...c6, all of which 2...Bg5 are either dubious or bad. It is fine against 1...d5 because the d5-pawn blocks any Qa5 tricks, like 1.d4 d6 2.Bg5 c6 3.e3?? Qa5+ -+

Something against 1...c5

So for me, it is:

The entire Trompowsky Complex

The entire Levitsky Complex

The French (1...e6 2.e4)

The Exchange Slav (1...c6 2.c4 d5 3.cxd5 cxd5)

The Modern (1.d4 g6 2.e4)

The King's Indian and Old Indian (1.d4 d6 2.e4 Nf6 3.f3 and now 3...g6 4.c4 or 3...e5 4.d5)

Closed Benoni/Schmidt Benoni - 1.d4 c5 2.d5 and now 2...e5 3.e4 d6 4.Nc3 or 2...d6 3.e4 Nf6 3.Nc3.

For Black, he only needs to know one of those lines. A single variation of the Trompowsky, a single variation of the Levitsky, or a different response and know that, like I play 1...e6 against 1.d4, so either a French, Dutch, or some minor line like a London or Jobava. But then Black needs to know what to do against 2.Nf3, 2.c4, etc. So if you are say, a Kings Indian Player, you need an answer for the Trompowsky (but only 1), and also a response to the London, Barry Attack, Jobava, Veresov or Pirc, and the numerous variations of the Kings Indian Defense.

So neither Black nor White has more control over the opening. It's just "different combinations of garbage" that both sides must know.

Black has just as much control as White after 1.e4 e5.

White can play 2.f4, 2.Nc3, 2.Bc4, 2.d4, 2.Nf3

Against those, Black also has options

2.f4: 2...exf4, 2...Bc5, 2...d6, 2...d5

2.Nc3: 2...Nf6, 2...Nc6, and a few minor lines

2.Bc4: 2...Nf6, 2...Nc6, 2...c6, 2...d6

2.d4: Only good move is 2...exd4, but it is really strong, and Black has multiple options against all White third moves.

2.Nf3: Black has 2...d5, 2...f5, 2...d6, 2...Nf6 (MY Choice), and 2...Nc6

If Black plays 2...Nc6, yes, White has options, but so does Black.

3.Bc4 - Italian, Two Knights, Hungarian

3.Bb5 - Berlin, Schlieman, 3...a6, 3...d6, 3...g6, 3...Nge7, etc.

So even 1...e5 players, it isn't like White is in control. You will get no Ruy, Italian, or Scotch.

1.e4 e5 2.Nf3 Nf6 3.Bc4 Nxe4!

1.e4 e5 2.d4 exd4 3.Nf3, Scotch can still be avoided via 3...Bb4+.

Etc etc.

As others have already pointed out, I'd say it's about equal for both players.

The first move for each player isn't hard at all.

After that point, the game goes in a certain direction. If we're staying in theory, then the players are technically just taking turns exploring a specific variation of whatever opening/defense they're in.

The knowledge required is about balanced, no matter which color you're playing. But of course, some players will come to the board more prepared than others.

Well you have a bit of a rigid take on this.

A players repertoire is always going to vary in its depth and rigor based on the frequency of the line, and how testing it is. In the model you're describing the player doesn't do things like this, he prepares with equal rigor for every possible move that can be thrown at him. He just plays like a machine all the time. But that's not how the game actually works. It never works that way - not even at world championship level.

For example, consider 1. e4 e5 2. Nf3. Here you stated black can play the elephant gambit (2... d5) and the Latvian gambit (2... f5). But these 2 moves are played 4% of the time combined, and they're not great. If white knows the first 7 moves white can just play chess and rely on his initiative to get him through the game. White would never study these lines with the same rigor as his Ruy Lopez lines, or even his Philidor lines. 
Now... after 2... Nc6 white gets to choose the Italian, Ruy Lopez, Scotch, or even the Four Knights. All of those choices lead to much more complex, dynamic games, and will require alot more theory to deal with than something like the Latvian gambit or the Elephant gambit. They're massive openings, and white got to choose which of them to enter. White can focus all his effort on studying just 1 of these, and so this is good for him. White also has more initiative, which will mean he has more viable opportunities to enter novelties. White is making the impactful decisions, he's controlling the game.

The initiative and level of control the two sides have simply is not always equal in a given line. Maybe to you and your philosophy of the game this doesn't matter, but I don't see a good rational justification for that viewpoint. It probably works by pure brute force, but there are better ways to go about things.

For your argument on the sicilian - black chooses the mainline open sicilian, and there are 13 different sicilians black can choose from. These lines are all much sharper, more complex, and require alot more theory to get right than the anti-sicilians. If white wants to be serious he'll need something against all 13 of them. Against the anti-sicilians it's generally sufficient to learn the lines 7-8 moves deep and play chess, the line will follow general sicilian patterns... there are a few exceptions, the closed requires a little more work (though 2... e6 makes it easy), and the alapin does too. But otherwise - you mentioned wing gambits, the bowdler attack, the smith morra, the mengarini... all of that is easy to deal with. At most black may need to know the first 4-5 moves there, not more. 
Maybe if some of these moves were played more than 1-2% of the time black would be forced to take them a bit more seriously, but as it is... it's a matter of priorities. 
Now, in your example you cited the Najdorf, but that's the largest, most theoretical sicilian black can choose. It props up your point - the move a6 doesn't control the center or pressure anything, therefor white is free to play a wide variety of responses. On the other hand, something like the Lowenthall is very forcing, and whites options are alot more limited there. Likewise, something like the Taimanov is very rare - white just doesn't have time to study that line in the same depth as he does the Najdorf or Dragon.

Now... continuing with this logic, let's say you're white playing the Taimanov. By playing a rare line in the taimanov you can equalize the experience differential. Hence... I like this line for white -

You should be a politician as you do nothing but emphasize extremes.

In your e4 e5 argument, you site only 3 of the 5 responses I gave to 2.Nf3. The two weakest ones, and the one that gives White the most option on move 3 (not necessarily the most options overall). I love how you just ignore reality and think the Philidor and Petroff constitute an insignificant percentage. Go tell Anand or Caruana or Smyslov or Karpov OR ME that the Petroff is insignificant.

You also clearly have no idea how significant some anti-Sicilians have become, especially Bb5 lines, even more than the closed or Alapin.

As far as the lowenthal, the Lowenthal is to the Najdorf as the Latvian is to the Petroff or 2...Nc6. Borderline trash, though not officially refuted, but if White knows, as you put it, the first 7 or 8 moves, White will have a significant advantage. And many of the lines previously thought to be minor have significantly gained popularity and hence dense theory, like the Sveshnikov.

And I am well aware that many opening variations occur more than others, but you still have to know them. Your opponent's rating often plays a factor too. You are more likely to face the open Sicilian against a 2200 than against a 1600. As a player that has played the French for 30 years, you get a lot of 3.Nc3 and some 3.Nd2 against masters and correspondence players, whereas against 1600 to 1900 players, a LOT of Exchange and Milner-Barry Gambit, both of which are far easier to defend than 3.Nc3. But you still have to know them! And there are masters that play the exchange. One even wrote a book geared for White, The Exchange French Come to Life, or some title similar to that.

But the fact that you only mention edge cases and ultra main lines distorts your point. Try playing against the Petroff unprepared. It is not a "dull, easy draw".

The extreme demonstrates the point clearly for those who didn't figure it out themselves and called into question what I said, but yes, black has his opportunities to diverge - and by playing the Petroff.. you're taking control of the game. That's a large part of what makes it good - if I was an e4e5 player as black I'd probably play the Petroff. However, if we compare whites major opportunites to take the reigns - Ruy, Italian, Vienna, Bishops opening, Scotch, Four knights - vs. blacks - Petroff.... what else? Philidor is not really in the same category here. White has alot more serious opportunities to take charge. And white has initiative throughout e4e5 - even within a line - let's take the Italian: white typically initiates the tactics. It's just the basic concept of initiative we're talking about - it's kind of like a dance, you have one person leading, and another person following. They're both moving in sync, but one person is making most of the decisions, one person has the initiative.

Now... you're correct that by playing something like the Petroff - which is a serious line - black might take back control of the game. However, for black to do this he first must understand the need, he must understand why he's not in charge otherwise, and what this entails...

Sure in the ideal world you should know all the lines, that'd be nice, but if a line isn't very serious / not common... it'll be easier to learn and you have better things to do.

Btw - I know very well how serious the Rossolimo is, hence I don't play the old sicilian. Very simple solution to that one.

Actually, in the Vienna, Black can often dictate the game in certain lines.

After 1.e4 e5 2.Nc3, the first option Black has is to allow the Vienna Gambit and play 2...Nc6, against which 3.f4 leads to a very wild game.

Black can also play 2...Nf6, entering calmer waters. Now the 3.g3 Vienna is more of a quiet game that Black pretty much as to follow White's lead. And 3.f4 only really sees 3...d5 as a good response. But the 3.Bc4 option, which is often viewed as the hardest to defend, Black again has a major choice. Do you prefer a wild game with 3...Nxe4, the Frankenstein-Dracula? Or more calmer waters with 3...Nc6?

A trend you may be noticing is that not all, but the vast majority of lines where one side has very little choice in the matter, the line he is defending against tends to be innocuous. 3.Bc4 is clearly more dangerous than 3.g3 in the Vienna. 3.d4 and 3.Bb5 are far more serious third moves for White than 3.d3 after 1.e4 c5 2.Nf3 Nc6. 1.g4 is weaker than 1.d4. So "dictating the opening" is not often (won't say never) a good thing.

Take your pick. Call the shots or have the better position.

The control I'm referring to is mainly mental. Initiative or "eval", as far as I'm concerned, is valuable in that it usually brings with it opportunity to diverge into novelty and take mental control over the game. The vienna has some of the best winrates for white in e4/e5, and by all appearances it's precisely due to white taking control in the way I'm describing. The line you mentioned as being bad, g3, is scoring better for white than the Ruy Lopez and Italian. You're arguing against a good amount of hard evidence here. It shouldn't be such a surprise when black is drilling the Ruy / Italian lines incessantly and playing them multiple times every day. I think you greatly overestimate just how much the 0.1 objective eval edge actually matters for the games outcome. If you actually study a line 10-12 moves in depth and your opponent doesn't know the line that well.... you will find an advantage in that line, I guarantee you. Such is the power of engines and preparation. But to each their own style.

On the other hand... your very own petrof is a perfect example of a line that takes control, and is also great objectively. Lines like this exist, and you should look for them. The QGA is another example... Taimanov is another. There are no anti-sicilians against 2...e6 which are seriously intimidating.

I never said g3 is bad. I said innocuous. Easy for Black to equalize.

The Albin Countergambit is bad. The Bongcloud is bad.

The Exchange French, Colle Koltanowski, g3-vienna, c3-Sicilian, 1.e4 e5 2.d3, they are innocuous openings.

I should also say that the +0.1 eval in the opening is bogus. Computers are inept when it comes to evaluating positions. You could play a line of the Kings Indian, be on move 12, and artificial intelligence will say it is +1.2. Then you have it plays what it evaluates as the best moves for both sides, and by move 25, it is +0.3 to +0.5.

Computers are great at finding forcing sequences, and evaluating late middlegame positions. It is horrible with Openings and Endgames. That is why you must load a 7-piece tablebase and an opening book for it to properly assess openings and endings, and not say 5 moves of the Kings Indian is +1.2 or KRN vs KR is +3 when in fact, the former is more like +0.4 and the latter, barring an immediate tactic at the point the position was reached, it is 0.00.

Depends on opening. For example, e4 has many good responses. d4 (especially the london) does not, so really just depends

As super GMs wage monster battles with the London system at key points in tournaments ... this statement ignores the practical side of chess.

There are theoretical advantage and then there practical advantages... Yes... on a theoretical standpoint something like the Exchange French might be innocuous, however .. with OTB, all it takes is one serious novelty or i.e. a game plan to gain a practical advantage. I'm very familare with The Colle Koltanowski There are lines in the Anti-Colle where White takes an advantage... maybe you don't know them, maybe you do) yea, it might be a small advantage, a pull in the middle game etc .. None the less it's an advantage. The main lines... with Nc6 and 8.e4 with black playing 10...h6 was the the refutation (if you could call it that) . No longer is that the main line. 8 b4 is.... leading to totally different positions where White plays for an opening advantage. All other main lines ...Nbd7, ...b6 ....c4 offers white chances to play for an advantage. (and Black too) Against the QID.. the standard setup, with playing the queens knight to f1 and working it up the kingside while pushing e4 offers chances. The Grunfeld ... (just one idea here) lines with an early b4 and playing the DSB to the queen side offers a fight. The KID... there are multiple lines all out of the standard main lines of a KID (like playing the French side of the KIA) that can... give one a practical. One call always play g3 main lines against the KID and the Grunfled. Last.... my favorite when Black plays a quick ...c5 after playing 1...d4. with 2.Nf3..... it often allows a Revered Queens gambit, either a QGA, some form of Semi Slav.. a reverse Tarrash etc... with move. Black might be equal in some of these ... but are they familiar with the positions?

Practical vs Theoretical

Granted these words might come with a grain of salt from a lowly 1400 elo player (1850 on lichess,,, Elo wise means the same here ) but I'm honest... I don't play with a data base opened, or with a VPN ... something I believe 1 out of 3 internet players are doing.