In my experience, rel="canonical" is very useful for the very reason that if the content gets served from a different URL, say an archival page or a misconfigur

In my experience, rel="canonical" is very useful for the very reason that if the content gets served from a different URL, say an archival page or a misconfigured server, you don't end up with two versions of the page being indexed by search engines. I try to add it to every page, now, so long as I can determine the page's canonical URL. I've found that it saves me a lot of future pain, when having to deal with duplicates that I may not even be able to control.