I have a similar experience with GIS programming questions. When I run into a problem with a specialized GIS library that I use, I can (in theory) post on StackOverflow since I'm trying to resolve a programming question. However, I almost always choose to post on the GIS StackExchange because I know I'll get the answers from more focused experts, more quickly. And even though I provide answers to GIS questions that I find on StackOverflow, I'll often also include a suggestion that the OP consider using the more focused GIS.SE site for future problems -- I know that it's a responsive, active site, whereas StackOverflow doesn't have nearly as many GIS professionals answering questions.
I think that this is an issue that can be monitored long-term. Ideally, we have a segment of users who are able to answer woodworking questions reliably and promptly, in which case it would be fine to keep such Questions on site. It is a type of craft, after all.
We'll need to consider factors like:
- How many answers do such questions get? (One? Two? Five?)
- How quickly does an answer come? (Hours? Days? Months?)
- Are answers thorough and detailed, or speculative and brief?
If A&C isn't providing thorough and (reasonably) prompt answers, then it isn't an ideal community for woodworking enthusiasts and they might be better served elsewhere. Migration isn't necessary, but commenting to let them know of the alternative resource seems appropriate. (Not really necessary in the private beta, since our woodworkers all seem quite aware of their options, but something to consider as we grow and develop.)