7% completion rate, and the users are overwhelmingly quite well-educated already... calling into question whether MOOCs can reduce inequality at all.

These numbers are surprising, but the fact that MOOCs mostly benefit the well-educated might be a necessary first step toward more widespread adoption. Early adopters tend to be well-educated.

I was surprised too by that low completion rate, especially since 2 of the 3 schools are for profit, I guess implying the students had paid for their courses; which usually would mean they are committed to the class.