This very interesting article by Jason Potts of RMIT on "Why MOOCs will fail" seems to confirm what I have for a while suspected to be true. He makes two main points, firstly; that young people go to university to get a mate, and secondly; to send signals to potential employers. They can get neither of these from MOOCs (or possibly all online courses for that matter) and that is why they will fail.
The first point sounds like a joke, but I do get the impression (from my own kids if not elsewhere), that higher education is a pleasant life experience, that they feel entitled to, and the potential to meet romantic partners is no doubt part of that. He does make the interesting point that a very substantial portion of the economic return achieved by attending college is due to marrying a partner of similar social status and income potential. I'm not sure that this is explicitly in the minds of young people who want to go to college, but it may well be hidden in some subconscious evolutionary psychological urge to meet "people like us" in a romantic way.
The second point I think, is much more important. Both parents and kids do think explicitly about how employers will view the education they pay for and receive respectively. This may be why they choose more prestigious institutions even though there may be no evidence that the quality of learning is any better. From my rudimentary knowledge of economics, I believe that this is a form of signalling. The peacock has no great use for his flamboyant tail, but the tail does signal to a mate that he is in good health and has excess nutrition that he can waste on growing and carrying around the tail (perhaps like a first year with an iPhone and Beats headphones). Going to a good college signals potential employers that (i) you had the smarts and/or work ethic to get in, (ii) you had the smarts/work ethic to get through (and perhaps surviving bad teaching as a bonus if the employer is aware of it), and (iii) you come from the right type of background (who are smart and/or value education). If this is a very reliable signal, it saves them a lot of effort in recruitment and the candidate pays for it.
So even if I were to suggest that online work-based learning is superior to traditional campus based education, it's not going to disappear very quickly and growth will most likely be slower than we zealots would like. I don't think MOOCs or online learning will fail. They will succeed for other reasons. We'll just have to live with this problem and work around it.