In general, when there's important functionality missing in the client, I think it's at least worth considering whether to add it in the core. For example, if other (i.e. non-Python) REST clients might want the same functionality, let's push it into the core.
It's fine to add a workaround in mailman.client first, as long as the API doesn't change once the core adds functionality. In that case, you'll be able to just change the implementation to the more efficient, core version, without code using the client library having to change.
When that's the case, it's fine to add a mailman bugtask and a mailman3 tag.