@Catija Well, that is probably true for most of the feature-requests posted on MSE. There is so many of them that it's not possible to implement all of them. Probably at least some of the more popular requests get noticed by SE at some point.

There are also quite a few old feature request which are quite highly upvoted, but not implemented. If I had to pick one that I would really like to see: Allow non-bumping minor edits, but review them on /review.

This was discussed also on MathOverflow Meta: Minor edits, subject to review. And apparently Scott Morrison talked about this with some Stack Exchange employees: chat.stackexchange.com/transcript/10243/2013/9/9 (Although it is not clear from the chat message whether it was with CMs or with developers.)

Maybe I expressed myself rather clumsily, but my point was: Even if the likelihood that a feature request is implemented is rather low in general, it is probably still worth at least to give it a try. And to do that, the best way is to follow the proper protocol and post a feature request either on local meta or on Meta Stack Exchange.

I'd guess that many users around here have similar viewpoint. (And I consider this a reasonable thing to do.)

This might be one of the reasons for rather low participation on meta.

And certainly, it is nicer if user spend their time by spreading knowledge than by discussions on meta, retagging, editing, etc.

On the other hand, I am not sure whether the site can really function well without at least some community moderation and at least some involvement from moderators.

This reminded me a bit of an older discussion I had with quid, I will link to message which is most closely related to this: chat.stackexchange.com/transcript/message/40741161#40741161 It was part of a longer MO-related exchange in chat: chat.stackexchange.com/rooms/19138/conversation/…