I'm not sure how long svn's life will be, but I agree that it was a useful transition. Moving our projects from cvs to svn was trivial, and we waited long enough for the end-user tools to have the features we needed. Basically, there was no reason to stick with cvs, at a time when it was not yet clear which dvcs we should choose. For many projects, it's still not clear which dvcs would be the best for that project.
On a personal project, I am now using bzr as the front end to an svn repository. Very cool. Yet another incremental step. I have been a strong dvcs advocate for a few years now, but a sudden complete switch is not always possible or smart.