I usually have a young squad so my players usually 'peak' quickly. But I also sell players that don't have the potential to become tutors in the future, that way I get a high price for players that have been very stable but are getting inconsistent. I don't like inconsistent players in my team, I sold one of the games top strikers Vaclav Kadlec for 20m because he was inconsistent.
I don't know if you like the Roy Hodgson style (using old and experienced players that really don't know football anymore and are slower than Giggs) or if you like using young players. I've seen how many people say you should develop your best young players, I usually keep them in my squad until I sell them (I prefer not loaning them), but if I don't see them becoming my star players but a backup player or even not good enough for my squad I loan them out. If I get a player from my own academy that has the potential to become decent/good EPL player I keep him, if that would be his scout report I wouldn't buy him.
Short Version:
It depends on your style of management, do you like to have young players that you can get profit from at the age of 24-28 or let them develop slowly and let them 'peak' at the age of 24-28.
EDIT:
It also depends on if you already have a star player in the team in the player's position. If I have a world class player in let's say the DR position I wouldn't let the young player play as much as I would if the main DR would be a decent or good player. But I usually let them play all of my League Cup and FA Cup games, I also put them in late in games. I'm currently introducing Martin Kelly to my first team but Glen Johnson is playing really well but I play Kelly late in games when Glen Johnson is getting tired and against weaker teams (because my first team players sometimes get complacent :S) and he usually has between 7.2+ in full games and between 6.8-7+ when he comes on with 30 minutes left.