My perspective in the earlier Meta post herein the earlier Meta post here was that a question about modelling without computational aspects would be welcome at Math.SE or MathOverflow (or perhaps a discipline specific site such as Biology, Physics, or CrossValidated).
My notion of computation is broader than just computer implementation, and includes discrete as well as continuous math. In that sense I expect many modelling questions to be on topic here. Whether you'd find a "deep bench" here waiting to address all of them is open to doubt, but in my experience the community is more supportive than antagonistic about scope issues.
I'll add some thoughts to the Area 51 discussion.