I enjoyed this:
Course completion rates, often seen as a bellwether for MOOCs, can be misleading and may at times be counterproductive indicators of the impact and potential of open online courses.
@vanessa you said you’ve read 1 of the working papers, is it worth it to discuss on a community call?
I read several of the working papers, in fact. Justin Reich at Harvard is heading up the research initiative at Harvard X. Maybe we could have him come talk to us? @bekka do you have his contact info?
I think that is the wrong conclusion. The conclusion I would draw from the data is that the course format doesn’t work for most learners, especially online learners.
I had to sign up to see their demo course 
They support open and content is release under CC-BY-NC-SA and then they put a signup in the way. It’s like queueing to walk down the street.
The early MITx I remember was just the course materials and assignments. It seems a bit more complete now - content, discussion, a wiki announcement and two dashboards. Does our current courses on p2pu.org significantly differ from this? I guess the Mechanical MOOC differs a from a tech perspective.
And when we move to something new, will that differ? Or will we draw the same conclusion from that - “that the course format doesn’t work for most learners, especially online learners”?