It's just a matter of semantics isn't it? I mean lets take Lukaku as an example. Say we buy him for £20m with a buy-back clause saying Chelsea can buy him back for £30m in 3 years time. On the plus side, they'd save on his wages, so maybe save £3m a year. In this scenario they pretty much break even.
Or, they could have the same player, and loan them out, probably earning them a couple of million a year plus potentially having the renting club contribute to his salary (if not pay it all), so potentially earning the club perhaps £5m a year. In this scenario they earn a few million each year
Scenario 1 would give Chelsea a big lump sum to re-invest somewhere else right away, plus a bit of money to spend on wages somewhere else each season. If Lukaku then developed into a Costa type superstar, then they'd still probably save money on what it would cost to sign someone of that stature.
Scenario 2 would see them lose out on the initial fee, but gain a larger sum each year and be financially better off at the end of the period.
So, it would seem that the only way a buy-back thing would be tried is if the selling club don't really think the player will make it, but want to hedge their bets. We've seen Barca do this a bit lately with the likes of Bojan and it's pretty clear they don't have a whole lot of faith in him.