That's a great idea, Dilaton!

Hey Paul, do you mind if I repost your 'answer' here to tea.mathoverflow.net ? Comments can also be moved. (Yes, it is another site to log into, unfortunately -- and one does need to register and sign in -- but I think it's more or less instantaneous, and it's a damned sight (or 'damned site') better than these comment boxes; I think it would be easier to have a civilized discussion there.

@PaulTaylor (1) Please go right ahead and invite outsiders to answer questions! (BTW, I was a little sad that the question about Bourbaki and category theory was closed, although there were some issues people had with the formulation (arguably not deal-breakers, IMO). (2) I wish you would at least be willing to try the mechanism, rather than reject it out of hand.

Dilaton: requests for reopens is semi-efficient IMO (it is indeed useful). Maybe people like Paul Taylor should be using it more?: meta.mathoverflow.net/questions/223/requests-for-reopen-vote‌​s As for the other question: I'm not quite sure what you have in mind, but it might be worth posting as a separate question.

I agree that the "research-level" mantra is somewhat overdone and somewhat misleading, especially when applied to questions that could come up in graduate school classrooms. (I think it's more understandable when applied to questions about grammar school mathematics).

still very much an ongoing experiment (and we welcome dissenting voices), I think "we" definitely don't want to go the unmoderated route of math.sci. So some careful oversight and moderation seems to be a good idea. Granted, it's a tricky balancing act, because we also want to attract good people from outside the old "core areas" as well, and we don't want to turn off the graduate students who are our future colleagues. Still, I disagree with Paul's suggestion that all but the outright obnoxious questions stay open (and recommend visiting MSE if that's what you think you want).

trafficked, and it's fast: as soon as a question appears, a swarm of mathematicians comes in (often with essentially identical answers). It's actually a very nice site, but I think we're trying to establish a different sort of functionality here. Many mathematicians like it here for the stimulation of the relatively high level, and don't want to wade through a bunch of relatively elementary questions to pan for the nuggets of the good stuff. Indeed, attrition of the high-level mathematicians we seek to attract is a recurrent worry (especially since the merger with SE). While MO is (cont.)

Paul has asked me to reply. I hate these little comment boxes, and may request porting the discussion over to tea.mathoverflow.net which in my view carries some advantages over this set-up. Anyway, I think Mathematics StackExchange might better fit the bill for Paul's desiderata: there too mathematicians from a wide variety of areas congregate, and accept questions from basically all levels. (They close in case of spam or offensive posts, or duplicate posts, and there are some others.) The atmosphere there is markedly different from the one here: it's much more heavily (cont.)

Sudhir: I understand. Of course you're welcome to try at MO -- we'll take a look -- but receiving an insufficient answer at MSE is not by itself a sufficient reason for keeping it open at MO.

Well, it's good that you ask, because it seems as if a lot of people suppose that if a question hasn't been answered at MSE, then it's time to ask it at MO. In many cases that's just not so, and I wish that more people would look around MO a little first, to get a sense of the place and see if MO would be a good fit for their questions before they ask.

(Lots of stuff redacted, very much in keeping with the spirit of the post. :-) But my final summary statement was:) The rule of thumb might be: delete your own comment if you yourself judge it is no longer useful, and don't worry about the rest.