You do pose some very critical questions here.
First , you must know that learning to attack(or anything else) is not something that it can be learned from one day to another.It needs serious study and effort.
One thing that certainly helps is to take well analysed instructive games (obviously you need good books) and replay the position from which the advantage begins , trying to win against your opponent(your opponent must be a much better player).You can start from a point where the advantage is winning.Of course It would be pointless to start from a point in which the winning side has forced mate in 3.You have to choose a position from which the advantage , no matter how big it is , needs some kind of technique to be exploited.If you don't win(your opponent won't follow the book , he will play what he likes) , the analysis can show you the mistakes and you can try again.If you win, you can also try again , this time from an earlier point where the advantage of the winning side is big but not winning.The same game can be played from 3 or 4 different points which will make you understand all the phases of the advantage, from the correct positioning of the pieces to the final exploitation.Knowing what you have to do from the analysis , comparing with what you did and determining of what you should do, will help you understand better certain type of positions.
I don't know if that helps but we saw impressive results in my chess-club with that method.
After trying the computer workout "development" a few times I have found that I am completely unable to turn a lead in development into anything meaningful. The computer closes up the position and with an almost complete absense of space manages to undermine my pawns and make it's position better. What I would like to see is a course on attacking purely defensive setup's, as the genre of "attacking" tends to fall under either grandprix attacking or going against a setup with fixed weakpoints.
In my case when I set up an attack against the computer opponent it removes the sting from my attack with a pawn move or two and suddenly all I've accomplished is an easily underminable pawn chain weakening my position with no solid attacking chances. I suppose a big part of my question is what are the inherent weaknesses of a structure that isn't changing from the starting position? Obviously this wouldn't occur in a real game, but clearly my ability to determine what a weakpoint is isn't that sharp.
So what exactly do I do? The computer is obviously not going to give me an advantage by playing aggressively when there is not a good way to immediately. Surely a lead in development means something though -- the question is how to approach making something out of it.
side note, the position from the "development" workout is a bit ridiculous just on the grounds that with black moving Nf6-g8-f6 or what have you leaves white with only arbitrary ways to develop anyhow.
double side note, I don't like watching a line of best moves only to be shown
"You did not play accurately. You slipped from a better position to an equal one."
because that is just silly.