IMO, its less useful than it sounds. Asking to make a step-by-step guide implies that there is always a right choice.
Trying to win at higher difficulties, IMO, involves a general understanding of the benefits and hindrances of each tech. By making a step-by-step guide, you are implying that there is a "right choice" at a certain point. I think it would breed players who follow something mechanistically, as opposed to understanding how it works. Rather than developing the sense of strategy that lets you win at high levels, you would be teaching people how to respond in a specific situation.
For example, @ "turn 1000, Go to war."
That may be the correct call in a specific game, but is incorrect in general.
If you copy someone else, you'll lack the understanding you need to win most games.
My recommendation for getting better would involve two types of games:
1) The all random game. Play a game on a difficulty that you can win, but set everything on random. Therefore, you have to work with what you get; improvise. You cant go in with some cheese strat (Quecha rush!). You'll be forced to learn how to use whatever advantages you have.
2) Play a game with a difficulty that you can win, but decide beforehand what you're going to try to learn. For example, make a game in which you want to use a specialist economy. Or a game where you want religious dominance. Play several games with the same settings, just with a different strategy/civ. Therefore, you can compare/contrast and learn from your own experience.
A final caveat: Some people have developed specific strategies that will let you win. Pre-2.08, things like the CS slingshot to macemen, the quecha rush on diety, or great wall -> pyramid were all "tricks" that could allow you to win at high difficulties. IMO, these dont really improve your skill. Granted, the person who developed them is very clever.
Hope it helps!