There will be rare exceptions (what they mean to the city, organization, generational type guy) but for the most part, I think getting rid of basically any player before they hit 30 is the way to go.
Sometimes, that means trading them at 27, sometimes it means 29. Situations and who the player is dictate all of those things but my basic rule of thumb would be to get rid of them by the time they hit 30.
I'm under the assumption that this is just a giant joke. Ripken is not even a good AA player. He would have never had been signed if not for his name and would have been released five times already.
He's not even a priority play at Norfolk. He's just an extra guy to give a guy a day off or cover down for an injury.
His slash line over his minor league career is .242/.281/.332/.613 in 1480 PAs. He's been given more than 295 PAs in a season twice.
Once in Washington where he slashed .201/.241/.254/.495 in 399 PAs as a 22-year old between NYPenn and Sally League (A-) and then with Delmarva (A-) at as a 24-year old slashing .244/.281/.327/.608 in 420 PAs.
Unless you just want to see a side show where a 3rd Ripken plays in the major league with the Orioles, there is not right to even be having a conversation about him being promoted to the Orioles. It would be a travesty to every hard working long time minor leaguer who actually have had a good minor league career if that were to happen.
There is likely a statistical answer to this and that answer is likely to have something to do with the profile of the player (e.g., velocity, age, LD% or whatever) and the parameters of what you consider successful.
Alex Wells is a great example. Very successful at AA but not the profile of ML success. Not impossible, but not the profile without velocity. It happens with hitters too. I forget his name, but we had a guy who won the triple crown (I think) in Bowie about 10 years ago. He was old, which is part of his profile, and likely had other profile things because he wasn't really a prospect even at that time.
Short answer is that success isn't a bad thing, but it's not a guarantee. Alternatively, it's probably hard to play at any level without success and then go to the majors and have success unless you're talking about very small sample sizes.