Would this be a better way to track my progress?

I'm looking at ways to set goals for myself, but I'd like to get away from rating somewhat. I'd like to change the way I think from "I need to win this game." or "I need to get my rating to 3500. (that might be a slight exageration)" to "I need to find the best move in every position."  The question is, how to track that over time.

I had an idea that I'm sure a whole bunch of people had, and I don't see anyone talking about it, so I'm assuming it won't work.  I'm just not sure why so I'd thought I'd ask.

Would it be a fairly accurate way to track my progress to take each game and "score" it against the computer, with the score each move being the difference between my move and the engines move, say always at a depth of 20 ply? The average of the total difference per move would be the game score.  I know engines aren't perfect and don't always find the best move either, but would it be a more accurate way of tracking my progress than worrying about rating?

 I'm a programmer, it would be pretty easy to build an app that would do this automatically and track the scores over time.  All of this would be done after I've reviewed the game without an engine.  I could also run games from others through to see where my goals should be (i.e. players rated 1600 average -.37 but I'm at -3.7). 

I don’t think that method would work. Let’s say that you’re playing a game where you’re ahead by a rook and you can sacrifice the exchange for a very easy win in the endgame. The computer might rank that simplification as being rated a lot less than some 10 move combination it had calculated out, even though both would lead to a win. That’s just one example of the computer punishing you not for making a bad move, but for making a human move.

A non-rating way to track your progress might be to keep track of why you lost your games. For instance, maybe you go from losing due to 3 move combinations to 4 move combinations. After studying a lot of tactics your realize that instead of losing due to 4 move combinations, you’re botching the endgame. After spending time studying the endgame you notice that your losses tend to come from games where you don’t activate your rooks early enough. And so on and so on.

That makes sense.  Thank you very much for the input.

You could also submit your long games to a coach or on the analysis forum.

The computer method you're describing is okay for very strong players, but at amateur level, it doesn't make the difference between a huge blunder losing a piece immediately and a small inaccuracy that may lead to the same loss, provided your opponent finds a fantastic 6 moves deep combo... It's also not very good to assess positional mistakes, that the human eye catches much better.