Not only do I love working for ENotes, I learn about possible questions my students may have regarding texts we cover based upon the questions posed by other students. I find that these posted questions allow me to address issues that I see on the site well before the issue arises in my own classroom.
As for the movement of "great discussion posts," I feel like the posts change with the times. As issues arise, posts are created. Sometimes the information remains relevant; other times, the information becomes forgotten. The great thing is that there is no limit on the times a post can be made. As posters define issues, editors can respond to them with old, new, or possible suggestions. Editors learn as they go, same as everyone else. Posts which are revisited can be "dug up" from the history or begun anew.
I really enjoy looking at some of the discussion posts on the literature side for some of the novels and poetry units that I teach. Those are great for creating warm-up questions or for literature circles, even for open class discussion.
This answers some of my initial curiosity. Of course, I was not suggesting that measures like these haven't yet been taken in any way--just that I think discussions have incredible potential as teaching tools and wanting to explore those possibilities further.
I hear where you're coming from. I think eNotes does a better than decent job of organizing a rather vast quantity of material on the site, making things searchable and orderly. Thinking about how to make it even better can't hurt though.
There is a "great discussions" sections of eNotes that seems to be a selection of discussions picked by the eNotes staff. This would seem to address the question of how to present an archive of high quality discussions, at least to some extent.
Most of the discussions that make the "great discussions" section receive many posts before going to that section then receive many more upon being moved there. The quantity of responses seems like it can be a good indication of a thread's quality. Maybe weighting discussions according to the number of discussions as well as the five-star rating system could suffice to separate the wheat from the chaff, as you say.
Maybe this is already being done though. Each category/subject/topic already has a "popular discussion" tab in its discussion section.