I'm just starting to really understand and appreciate the dynamics of position and I think this is one of the textbook situations where position makes a big difference in how to play a hand.

So, what I'm trying to do is evaluate the outcomes given all possible holdings of the opponent:

1. If he's ahead, then 3-betting is obviously bad because villain ain't folding and hero's losing more.

2. If hero's ahead, then either (a) the flop 3 bet will allow villain to get away from the hand when hero'd probably prefer for him to keep putting money in the pot, or (b) for that narrow range of hands that we're ahead of but that we'd like to fold out (2 overcards), the flop 3 bet inflates the pot and makes a call by villain either correct or at least less of a mistake -- as compared, e.g., to calling the flop and donking the turn (and folding to a raise).

I think this approach applies to most situations where hero might already be behind and is OOP. Probably applies even more in a case like this where he's facing reverse implied odds.

But I could be way off here.