I would like to thank Todd (whom I know to some extent) for telling us how MathOverflow came to be set up and also the people he mentions (whom I don't know) for doing it.
I see it as an extremely valuable site because it is the only place so far as I am aware where mathematicians of all disciplines congregate. However, it is precisely for that reason that I think it worth expressing a dissenting view.
I would like to see an end to the practice of closing questions (unless they are actually spam, offensive, etc).
Painfully frequently I see the high reputation users closing questions that plainly come from beginning graduate students, thereby humiliating them. Remember that these are your future colleagues and Fields Medalists. Do you behave in this way towards your own advisees when you meet them face-to-face in your own universities? Do you not have the imagination to turn naive questions into teaching opportunities?
Closing questions on the grounds that they are textbook ones is necessarily inconsistent. Quite difficult points of say number theory get closed when outsiders might have appreciated some explanation. Questions in logic that are comparatively easier (in my opinion) don't get closed; rather we have users like Joel David Hamkins who take the trouble to explain things, and it's a good thing they do.
Then there are the extra-mathematical questions, such as those about history and methodology.
Rhett Butler gave some excellent factual answers to numerous historical questions but somehow he no longer participates, probably on account of the (in my view hostile) reception he's gotten.
There was a question about Bourbaki and categories where I could have found people outside the site to give authoritative answers. But, had I done so, they would have arrived to find the question closed, which is insulting to them.
Amir Asghari asked a question about rigour that led to a flamestorm. However, it was only after that was over that I myself read his profile carefully and saw that he was trying to understand the cognitive processes behind doing and learning mathematics.
The value of this site is as I say that it encompasses many disciplines. Its participants should recognise that and value the role of people from outside the old core areas.