Surely the fact that revisions per article is rising just as much as the number of articles is hopeful for the article quality. (Medians rather than just means would be good to know, of course.)
Hopefully it is that, rather than an indication of an increasing frequency of edit-warring.