controversial articles and neutrality problems
You're missing a key point still: "A substantial population"
The principle is not that we delegate I-P (or fringe science, or others) to a couple of users to sort out. We still use the classic crowdsource model, but the crowd is of users (from all topics and areas) that the community has agreed consistently edit to core principles and a high standard, including in dialog. Not "just a handful of moderators". Those users whom the community has not yet formally agreed edit and interact with a consistently high standard can still participate, but stick to the talk page.
In that editing environment, a user who is tendentious or greatly biased, or a subtle warrior, won't stand much chance. They, and any users who obstruct good editing, will rather visibly be bringing their own standing as a "senior editor" into question (which is a high barrier to regain if lost)... because the peer group of editors is other users who are acting to a high standard -- and recognize/expect the same.
(In a practical sense, if a quality mark takes time and effort to acquire, and is readily lost if abused, then those acquiring it will want to not lose it. Gaming the system is a lot harder in a group of high quality peers who know what the norms are and what is expected.)