Let's try to be civil this time around.
#1 Training - Don't you think this makes training really easy. There is no downside to trianing any more. Why do you think this change is needed? To help higher division teams to train? It's certainly not a problem in lower division (as the players can actually play effectivly on most training spots).
#2 Draft - You are going for realism here, but the fact is that some thing, although realistic, do not work well in a game. Lower divisions have their niche, it's training. If DI teams don't want to train their 18 yo's, then they have an option of buying a player, that has been trained for 3 seasons by a lower division team. This suggestion only adds moneytary value to higher division draft. Which is not good for the game.
#3 income - With this part I partly agree. Fan survey should reflect home and away games separately or it would take into account last 2 games. Last 10 games is too big of a history, it will add last season games aswell, for relegated teams. Relegation is already a big burden to shrug off. The simplest way to adress it is to raise the player salary floor. I also think teams should have a roster of atleast 8-12 (nba has 12 active list requirement with some exceptions) players. Having less will make you an automatic WO on game day. I also would also like to suggest minimum lineup of 8 players for any competitive game (cup can be a place for arguments). You can list minimum 5 players for scrimmage. Having a WO in a week, negates that weeks income as a fine (tv money, merc, arena income). Player transfer income does not get touched.