The problem I have with the OP system is that it doesn't take into account teams like Oregon State. As mentioned above, James Rogers came in as a 2 star and is now a respected top threat in the Pac-10. However, this is the norm at OSU.
I read this article recently which told Beaver fans what we already knew. It talks about how 7 players from Oregon State were drafted last year, 2nd highest of any NCAA team.
http://rivals.yahoo.com/ncaa/footbal...n=ncaaf,184458
Each of those players was relatively unrecruited, and yet OSU consistently looks for those diamonds in the rough that transfer into All Pac-10 talent.
Unlike the OP's idea, I think there should be a lot less assurance that a 5 star is going to pan out, and a lot more random 2 star, 1 star, no star players that make a huge jump at some point in there careers.
I would do this:
First of all, have incoming freshmen go through progression too.
Second, a lot more random jumps and falls (yes, players getting worse, mainly in awareness and whatnot) so a 2 or 3 star could randomly shoot up.
5 star - Usually has the physical attributes, like speed and what not, but can regress before the first year in various categories. Awareness. Catching (receivers). Accuracy (QBs). And make all of them mean something. Make it dangerous to start a receiver with only high speed, because he drops the ball a ton. Or is not able to get separation.
4 star - Even less sure. Again, usually having the tools, but may come in with a lot lower rating than expected, some might be studs. Some will wallow and slowly drop in rating throughout their 4 years.
3 star - Some guys that have the tools, but low awareness and whatnot (scouted) but some come in and have those rating shoot up (not many) some go way down (as if there confidence gets crushed) but most just be mediocre their first couple years, with a few making big jumps here and there. Have some guys that don't have the tools, but high intangibles and get better at the tools (not great) to make them respectable players.
2 star & 1 star - Still a few that surprise out of no where. A few here and there with good tools and terrible scouted intangibles but come in and are better than advertised.
Which brings me to another feature: Scouting. Not too in depth. I don't want to spend too much time on this, have a good automatic option that does a pretty good job of scouting on it's own, but that way you get a star ranking and some idea of their rating, but could be a ways off.
Finally, each school have a rating on how they develop talent and have the teams that have a very high rating get just a few more of those diamonds in the rough that jump up. Say the average percentage of 3 stars that make a jump is 2%, well for a great developing team (like an Oregon Sate) maybe 5% make a big jump.
Of course I also think we need dynamic ratings that change during the season. A lot of good ideas here, if EA would only decide to actually put some effort into improving the game.
Maybe it's not fair to lambast the EA devs like this because I've never worked on a game project of this magnitude (though I have worked on smaller games) but the proof is in the pudding. Some slight gameplay advancements and an interesting new feature don't cut it for me.