It is good to look back at the history of MOOCs and the influences and context that brought them forth. Martin's interview with George and Dave gives an excellent sense of the pedagogical challenges that were taken on and solved to a large extent in the early cMOOCs and are still confronting the xMOOCs today.
The lengthy report by McAuley et al being written 3 years ago and predating xMOOCs seems like ancient history and almost 2 MOOC generations old. Even Martin's MOOCs Inc post with the important link to Curt's May 2012 massive link bait seems to exhibit peeling MOOC paint. Curt admits that even back then he had too much reading, watching and listening to do to catch up on MOOC developments. So how much harder is it today?
Web Development. So can a MOOC compete with a traditionally lectured subject? You bet it can. In 7 weeks of the CS253 MOOC Steve Huffman and 1 TA covered about double the amount of material of my 12 weeks of lectures.
As it happens we both used Google App Engine for the practical work and this allowed direct comparison. I taught about 15 students and Steve at least 1,000 times that number. What I found most impressive was how App Engine allowed the used of automated assessment scripts so that students were able to receive direct and useful feedback with comments on their practical work 24/7. To moocify my own course would be very straightforward.
The drawbacks would be that my institution markets and prides itself on small class face-to-face teaching. Also I believe most of the weaker students would struggle after about 3 weeks of the MOOC pace, probably explaining the high dropout rates from which most MOOCs experience.
As far as I know the furthest my previous colleagues have come is to put up course lectures and special presentations on iTunesU, where I have to admit they receive acceptable viewing rates. A noticeable trend is that the shorter the 'lecture' the higher the viewing figures.