An important point is that academies, so far for MLS, have not been effective in producing cost-effective players. I happened across this report on the Prozone software website (I think its the software MLS uses to create the detailed statistics you find in game reports). There is a link to the full paper near the bottom of the story. The report is focused on their development of a player quality measure they call AREA. Not a focus of the report, but what jumped out at me was in Table 1. NCAA-background players are the 'best buys' for MLS currently. They tend to produce better quality by their measure, on average, than any of the other sources for players. Academy players have the lowest quality by their measure, on average. But, Academy players also tend to end up costing more than college players. They are quick to caution that the academy system in MLS is still developing and may yet provide better quality players than NCAA. But, the NCAA is currently the best source of players.