It seems to me that the Open Source community uses the term "project" rather loosely. PMI (the place for pointy-headed project managers) defines a project as having a beginning and an *end*.
NetBSD lists 5 goals on its website (http://www.netbsd.org/Goals/). It seems to me that NetBSD has succeeded on all 5 fronts. So have a number of other 'projects' succeeded at meeting NetBSD's goals.
As a past NetBSD-o-phile, I'm bummed to see things in the state they currently are. Fortunately, Charles has given the OSS community a great opportunity for self reflection.
I'm a great fan of diversity, but I think there might be value in a consolidation process. If everyone creates an indefinitely scoped project, I worry that we'll always be exploring side branches of an evolutionary tree. Put more bluntly, I think OSS will improve faster, and be more rewarding (especially for beginning coders) if the OSS community would set smaller, more discrete goals. Periodic pruning might keep OSS developers fresh and save them from burnout.
It may be that evolution will consolidate for us. The problem is that evolutionary development is slow. I'm not calling to shut down a bunch of 'projects', but maybe instead for individual developers to think smaller and reflect on their goals...