Zack Rosen just posted a comparison of Moodle and Sakai, based on some available online web- and project metrics (not an evaluation of the software itself). The comparison reads like something written by a Moodle supporter, intending to show how viable it is when compared with Sakai. It is viable, but language like "All signs point strongly towards Moodle kicking Sakai's butt and to the Mellon Foundation, Hewlett Foundation, and Sakai Partners wasting $6.6M" really isn't productive or conducive to an objective comparison.
Regardless, there is some surprising stuff in there - I didn't realize Moodle has 8,900 installed copies. I'd guess the majority of that are "tinkering" or "sandbox" copies - I have 2 instances of Moodle running here, but it's only being used once for students...
A more interesting "deployment" stat would have been number of campuses using each product with students. Number of campuses using each as their sole LMS? Primary? Secondary? Experimental? Average number of students per deployment. Number of administrators or server babysitters required per deployment.
The other interesting item, which should have been obvious in retrospect, is the funding differential. Sakai had a generous $2.2M round of seed funding from the Mellon and Hewlett foundations, with another $4.4M added by the core partners (likely a lot of that "in-kind"). Moodle is run on donations - at $12K/year - plus any time contributed to the open source code (which is likely rather substantial, yet uncounted).
The web metrics are essentially meaningless for comparing the projects, except to say that lots of people know about both.