Too many distributed key/value store projects?

I seem to hear about a new distributed non-relational data storage project every few weeks.  While I'll admit that all of them are pretty damn cool, most of the coolness is in their roadmaps.  Richard Jones posted a great overview of many of these open-source projects last week.  Given the fact that there are so many projects each with so much more to develop, and the fact that most of these projects are still early-stage, many will die off before they produce anything you can actually use.  I'd love to see some of these developers get together and start merging their efforts.  What do you think?