I would consider that an abuse of rel=canonical, and I'm rather surprised that Google seems to be honoring it. (All the results I see are from mirror sites without the rel=canonical link, whereas searching for the top-voted answer works fine.)
It would be one thing if the canonical link pointed to a "view all" page, but (AFAIK) StackOverflow doesn't have those. Here, the second page contains a substantial amount of content that the supposedly canonical page lacks.
Then again, I guess the net effect of doing this isn't so different from just putting noindex meta tags on the second and subsequent pages, except that all the PageRank for those pages is shifted to the "canonical" page instead of being diffused across outgoing links. So I guess I can see why Google might consider this tolerable, even if seems like doing a disservice to the readers.
(The proper thing to do for such pages, IMO, would be to use rel=prev and rel=next on the pagination links, which SO already seems to do. According to the page Su' linked to, Google takes these attributes into account and generally leads visitors to the first page, presumably unless they specifically search for something only found on the later pages.)
Update: Google seems to agree:
"While it's fine to set rel="canonical" from a component URL to a single view-all page, setting the canonical to the first page of a parameter-less sequence is considered improper usage. We make no promises to honor this implementation of rel="canonical"."