Thanks for the idea! The problem is not finding these sections; the problem is that, everywhere such a section exists, I need to dialogue with the author to clarify precisely what they wanted to say, and then I must ghostwrite a new version based on their clarifications. In other words, finding and identifying problems isn't the brain-breaking stuff; the tough stuff is deciphering it, the "search for meaning" in a linguistic no-man's land.

I think this is the misconception that a lot of outsiders to the profession have, and this is why they use AI instead: they assume the job is really nothing more than finding some simple mistakes, so they figure a machine can do it. The job does entail finding simple mistakes, but the bigger portion of the work revolves around the fact that human beings often struggle (and fail) to articulate what they actually mean, so it ends up coming out all jumbled, unclear, or even self-contradictory — whether in a research text or in daily life. ;)

I appreciate your suggestion, though!