This question is similar to the question "How to answer R questions" How to answer R questions but for math.
Hence, if people ask math question, we can
- Bounce them all to MO
- Answer all questions, even when it's clearly mathematic and not statistics.
Certainly the "poblem" won't as frequent as with R questions but 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.
I would say 2, with a possible redirection when we believe that a better answer will be gievn by the math community...
We already have an example of overlap (however I don't think this one is a problem) Introduction to statistics for mathematicians