Ok. I think I know what my problem is: if we can't reproduce the logical argument of a commentary, it might be useful as a historical source, but it doesn't (in itself) count as an answer that is backed up. For some questions, that might be the best we have, I suppose. But in general, we should be able to reproduce the work that a commentator did to arrive at their answer. If we can't, the answer fails to show it's work
even if the conclusion is correct. No commentator is authoritative enough to be above this standard. —
Jon Ericson 12 hours ago