The problem I expect a useful study of metaethics to solve is "How do we compellingly demonstrate that moral language (that is, our use of words like 'good,' 'bad,' 'right,' 'wrong,' 'should,' and 'ought') can be cashed out entirely in non-moral terms (e.g., in terms of expected and realized value of actions) with nothing important left over?"

Or, if we can't solve that one, then: "After we cash out moral language into non-moral terms as far as we can, what exactly is left over, and what is interesting about that stuff?"

It's not clear to me that this is entirely a task for evo psych, as what seems to happen is that regardless of what evo psych demonstrates, people who believe moral language cannot be cashed out in non-moral terms will simply deny that evo psych's claims are at all relevant.

Ditto for every field of study other than morality itself.

Of course, if I already believe that moral language can be cashed out in non-moral terms, there are several positions I can take:

• Yes, I believe this, but I can't compellingly demonstrate it.
• I believe this and can compellingly demonstrate it... this problem has been solved. If other people choose not to accept my compelling demonstration, well, that's not my problem.
• I believe this and don't consider compellingly demonstrating it to be a worthwhile use of my time.
• Etc.