I thought I should point out this post on the Statistical Analysis beta meta site; there is obviously a corresponding discussion to be had on this side. As David Speyer pointed out in this SBS post, a (small but noticeable) number of questions on MO really ought to be directed to statisticians; but as the meta.stats post points out, the boundary is quite fuzzy.

Of course if we don't want to be SE software chauvinists when referring questions elsewhere, there is also Metaoptimize. If you're wondering about the difference you can check out this post but I think the real answer is "these are very new sites and it remains to be seen how things will develop."

Sorry, I thought the meta site was publicly viewable since I could still see it after logging out. I guess there must have been cookies allowing that.

The post itself links to other currently private pages to set the context so I won't bother to quote it in its entirety. The title is "How do we handle overlap with mathoverflow", and the poster (Robin Girard) writes that if a "math question" is asked the options are

1. Bounce them all to MO

2. Answer all questions, even when it's clearly mathematic and not statistics.

and observes that "the frontier between math and statistic is not clear at all and it will be dificult to know and define if a question has 'too much math' or not." Robin's suggestion in his post is "I would say 2, with a possible redirection when we believe that a better answer will be gievn by the math community..."

The second post I linked to was "What advantages does this site have over Metaoptimize?", but I think there's no real point quoting it until the site enters public beta.

Perhaps it should be http://stats.stackexchange.com/ ?