What do you think of the idea of having a subforum in Release Candidates for "Next" with a locked, sticky thread where developers list out areas that need some testing focus?
If it helps, I'd certainly be willing to do that.
Not sure whether I would know what requires testing focus though... Many bugs are actually not in new features which get a lot of testing before they are committed, but caused by 'innocent' maintenance things (so the missing light - wasn't broken by maintenance on lights - that would have been tested before committing - but by restructuring terrain info storage).
So such a list would only be useful if it's finite - once I list every subsystem I touched with a commit, it ceases to be helpful. But to keep it finite might be problematic.
I think what is important to realize is that most developers probably have rather different use and test profiles of FG. I do 90% of my testing with AW and ALS on (simply because that's what I develop), shader quality at highest and check breakage of other rendering frameworks only comparatively briefly in the remaining time. Usually I just do a few minutes with the ufo at some benchmark locations (Caribbean, French Alps, South Africa, US Southwest) - I do 30+ minute flights in aircraft except the ufo very rarely and never any hour long airliner flights.
Obviously, testers which have a different use profile are more valuable then.
In general, I think I have the advantage that rendering tends to produce very visual bugs, so I tend to get notice soon. Though the disadvantage that they're highly architecture dependent, so knowing about a bug might still not help much.