I'm better at finding problems than at solving them :) But still:
Spreading talent out as evenly as possible would be a good start: while it can be disappointing that the best player in your pool is a superstar, is even more disappointing when the best player in your pool is a superstar (maybe with bad skills :P) and your neighbour league pool get 2 mvp and an hall of famer. I know that in real life there are very depth pools and very shitty pools, but it looks to me like one of these thing were trying to stick to realism make the game less enjoyable... trying to get a good pick to start from scratch after a bad season should be a legitimate strategic choice, but it's hard to do it when there is SO much luck involved.
Speaking of that, other thing i noticed is that a 5-balls prospect (talking about skills now, not potential) are too much different from each other in terms of skills and salary. While i understand that skills distribution in a player is randomatic (i suppose), so it can happen that you get a low value in crucial skill thus lowering the value of a good player, i've seen 5-balls player going from 4500 salary to 7500 salary - isn't it too much of a gap? once again, i just have to put the 5-balls first and pray to be lucky, cause the difference inbetween the 2 extremes is very deep. i'm not asking for 7k players everywhere, but it would be nice to have a better perception of a top player overall abilities.
in that regard, i love the addition of age and height - expecially height has always been another blind shot totally related to luck, and of course i hated it :) knowing that the player i'm trying to pick will be good for the kind of training i'm planning, and that i can improve him faster if he miss something, will make the choices more rational and less depending on coin-flips :)