A while back, Niels van Reijmersdal asked a question about the closing police. Since then, they've gotten the more negative title of the "closing mafia".
I've just gotten enough rep to help review the close queue, and while I've tended to close questions if it looks like the OP isn't being responsive to improving it, I've also tried nominating questions I voted to close for "reopening" if the OP does come back and make some attempt at improving them. I think this is the intent of being able to close questions as "unclear what you're asking" or "too broad".
However, my initial impressions are the regular close-voters tend to be resistant to reopening closed questions. See Why does it take a long time to spider an application? as a recent example--while it still isn't a great question, the OP has at least put some work into improving it and shown that they're responsive to feedback (I had left a comment about why I was voting to close, which I deleted when I voted to reopen after the OP's edit).
I'm concerned that this may discourage the very behaviors (and users) we want to encourage. We certainly get a lot of novice users that show up, ask a bad question, and then are never heard from again, so when a user demonstrates that they are willing to respond to feedback, shouldn't we give them the "benefit of the doubt" even if there's still additional room for improvement? Seems like if we just close the question and ignore their attempts to make improvements, we're telling them to go away and never come back.
I agree with the sentiment in the answers to Niels's question that leaving a comment when voting to close is good practice, but it seems like that's even more important if someone votes to reopen and you vote to leave closed, assuming there's been an intervening edit.