The first one starts with black up three pawns and is a tactic to show that grabbing that second pawn was wrong and the tactic evens the material and is also a winning position (by evaluation)
The second one is more ambiguous, though it's the only line that maintains the advantage. I'll ask about that one
Thank you for the answer.
Yeah, I understand that you win some material in the first one, but you still end up even material. In the end, it is winning but there is still work to be done. That is why, for me at least, this is not really great for a tactical puzzle, more for some lesson about king safety.
Usually the puzzles provide more concrete position in the end, like material advantage or a straight up win by checkmate.
I agree that the second one is more problematic, because you are an exchange down, even though it is obvious that you have a great compensation, and that this position is also winning and nobody would like to be in the shoes of the opponent.
In the past, I've noticed a few more that are somewhat like those 2, but only now I decided to post an example.
I think going from being behind and at least equalizing is a valid puzzle, in and of itself. Recognizing a blunder and capitalizing in it is important. There's likely a lot of similar puzzles, plus ones like forced stalemates and repetition is along the same vein.
As for stalemates and repetition puzzles, I completely agree. There is a sense that the solution is final, and in that sense it is similar to checkmates.
If this was a puzzle where after you gain back material, you get stalemate or 3 fold repetition by force, I wouldn't have an issue.
My issue is that this is the type of puzzle where you go for a win. And the win by the end of the puzzle is not really straightforward as most puzzles are.
It is more like when you have a game, and you are sure you are winning, but you don't yet see something concrete for the time being (like a checkmate or gain of material), and the tactical shot will come at some time if you don't allow counter play. This is my feeling of the end position of this puzzle.
I feel that puzzles should be, and they almost always are, more cut and dry than this.
But I respect your opinion. After all, the end position is winning and you won some material during the puzzle. Hope to hear some feedback on the second one at some point.
Realistic observations. And the problem will be bigger as the engines get stronger relative to human chess skills. For instance, lines now rated as "+0.34" may be rated as "+3.7" or "+M15" by the next generation of engines because they can read deeper with less effort. We see "negligible advantage" in the same position the engines see as "easy wins". We don't get better, but the engines do. The opposite will also happen. We see a strong attack worthy of a "+2.5" score but the engine sees an effective defense 10 moves down the line and only gives you "+0.7". Utimately, the perfect engines will score every position as either "0.00" or "+M47" or "-M93" which is of no help to any puzzler.
So it is a mathematical problem caused by the operating assumption that higher eval scores are more recognizable to humans than lower scores while the scores are calculated on the basis of engines playing engines!
What currently saves us from the situation getting more out of hand is that the engines won't demand you to choose in dualed lines (more than one winning move) and therefore often stop at seemingly random points handing you the "solved" freebie. While you breath a sigh of relief as you had no clue how to continue!
Until the design of puzzle engines is drastically revised - like including scales and models for human chess skills - the situation in general will get only worse with the engines getting further ahead of the humans!
This message is approved by the Union of AI-bots, the Galactic High Command of E.T. invaders and the Federation of Stakeholders in Quantum Computing. 
....
The second one is more ambiguous, though it's the only line that maintains the advantage. I'll ask about that one
This has been removed from rotation.
For instance chess Tempo is usually different. I am not saying there aren't bad puzzles there, but mostly you get something concrete material wise or checkmate wise by the end of a puzzle.
I have no problem with using engines when they are making these puzzles, but I feel they overdo it sometimes. Perhaps that is why some of these seem a bit artificial, even though I see they are from real games. In essence the end position has to be clear without looking at the engine, without a doubt. For the first puzzle it is clear to me, though still, it is equal material, and both players need to play more moves to really make it clear.
The second one is similar, but you've only won a pawn and you are still an exchange down. Sure, you can see that the compensation is there and that your side is most likely winning, but I feel that these kind of puzzles are missing the mark. Those are better like evaluation exercises where coaches ask which side is better and why, and so on.
Another thing I've noticed here, though I don't mind it that much, because at least you understand that you win in the end. Here, when there is let's say checkmate in 2 or 3, in many instances the puzzle stops after the first move. This is nitpicking, sure, but it will not hurt if they finish the line if the checkmate is a move or 2 away.
....
The second one is more ambiguous, though it's the only line that maintains the advantage. I'll ask about that one
This has been removed from rotation.
Appreciate it. Thank you for the information.
https://www.chess.com/puzzles/problem/996140
Here is another nonsensical one. Don't get me wrong, sure, this is winning, but this one move is not nearly enough to end the puzzle. It either needs more moves, or to not exist at all.
I am reporting them myself, but technically I shouldn't because the solution is kind of correct, it is just that by itself this is an incomplete puzzle.
I would like to show you two puzzles:
https://www.chess.com/puzzles/problem/986836
https://www.chess.com/puzzles/problem/982154
Now, don't get me wrong, by their end you get a winning position, but in my opinion I am not sure these are suitable as tactical puzzles.
Why? The aim of a puzzle is to either gain material advantage or to win by checkmate (or to find a draw if that is the best you can do).
These two don't do either.
For me, these 2 would be more suitable as a lesson in piece activity, king safety, in the first case the strength of a bishop pair perhaps and so on, but as tactical puzzles they lack the, should I say, finality one should expect from a puzzle.
I would report them, but it says that you should only report those puzzles where the answer is wrong or there are more correct answers, so it seems pointless to report them.
What do you think about those 2 examples? Are those 2 legitimate tactical puzzles or should tactical puzzles aim for more concrete wins.