How socially sticky pages are seems to be playing a role in new rankings, at least for longtail. If the site has RSS, how many subscribers is important. If the site has comments, how many on average is important. If the site is a forum, how many different posts by DIFFERENT authors is important.
Basically anything that can be used as a measuring stick had better show something besides non-use. If you never get comments DO something to get some or REMOVE the feature (including code footprint). Same with RSS features and other social interaction. If it's not used, improve upon that or remove it, you might be downranked because of non-use and it would be better to send no data than negative data anyway.
I base this partly on a forum I found that games Google search. It places 3 to 7 "similar threads" links with snippets at the bottom of each thread but has them appearing to be forum posters complete with different names, account age and they don't all show up at once... if you refresh the page another link/snippet shows up and another and another. The fake posts show up as real in Google search and the site's rankings have jumped recently, by a lot.
edit: I'm positive Google will catch onto this, I'm not advocating you do this (quite the opposite) but the behavior and its results are worth mention. It implies site USE is important. If there is a great article in the forest and nobody is around to read it... why talk about it?
A GREAT feature I've seen recently is with blog and forum comments allowing an extra field for visitors to post their twitter ID prominently. Every comment had one on this site, it made the site interaction more real and encouraged use a great deal (not to mention getting the pages re-tweeted like crazy).
[edited by: Sgt_Kickaxe at 6:41 pm (utc) on Jun 21, 2010]