Just to add to this, I don't think courses with significant online components are necessarily bad or evil or wrong. I just don't think that they're either going to be significant cost savers or significant revenue raisers. That's what I meant by saying the business model isn't there. Grifters and administrators fantasize about a course in a box, where you just plug it in and let it run forever while students continue to fork over 50 grand per year. It will never work that way. Such courses will still be labor intensive, and free self-directed learning options do and will continue to exist. If elite institutions open online enrollment to a wider variety of students, and include the credentialing aspect, they will degrade their brand.
What's lost in this discussion is that the cost per student per course for most professors, even relatively senior ones at relatively prestigious institutions, is relatively low. The large introductory courses MOOCs are imagined to replace really don't cost anything, even with a (relatively) highly paid full professor doing the teaching. When I taught at UC Irvine I earned a decent pay and had a decent course load. Over the course of the year I probably taught 500 students. Throw in a couple of TAs for the big auditorium courses and total instructional labor cost was probably $140 per student. Yes, plus benefits and other overhead. But the point is the cost of paying me was tiny relative to the tutition they were paying for those courses. There aren't cost savings here, because the costs are already really low (per student) for these kinds of courses. And the only way to have them be revenue raisers is to sell out the brand, which won't work either.