I sign up for this and that course on Coursera on a whim; but there are certain courses that I really intend on completing if possible; yet my real life does intrude. But these courses, while extremely good and extremely satisfying to complete, are meaningless to me. I own my own company and am not working toward some sort of degree(not that coursera courses add up to a degree). Nor do I need to impress the HR department.
But if I were a dedicated student who needed these course to complete some sort of degree which would lead to good jobs then I would select only those that I knew I had time to do, I would set aside time to do them, and I would not let a bit of irritation with the course cause me to drop it.
So to predict the future of real online education through coursera is mostly BS. But there are quite a few areas where I think coursera statistics could be interesting. For instance how many high-schoolers are completing fairly advanced University level courses? How many existing university students are taking the courses because their own university doesn't offer them or their local course sucks? And how many people are diligently going through Coursera because they don't have access to a University for whatever reason (money, weren't admitted, family issues, etc)?
Then there are the multi account issues. In many situations I will sign up for something like Coursera using an account like Poopy Von Poopy Pants; then realize that Coursera is cool, and sign up again for a real account while abandoning my "test" account. Lastly many people might have zero interest in interacting with the Coursera system and will sign up, grab the materials, and then complete the assignments on their own.
My first prediction for real courses, that divert from the traditional model substantive way (as in anyone can sign on for a reasonable cost without any complicated administrative crap) is that smarter, motivated kids in junior high and high-school will very quickly accumulate university level credits exceeding those for a first year or even meeting the requirements for a degree. This will eviscerate the high-schools as there is little reason to continue high-school if you are well on your way to completing a university degree. This can become a feedback loop where with many of the smartest most motivated kids gone the high-school experience will be dumbed down driving out the next wave of fairly smart, fairly motivated kids until all that is left will be the worst of worst, which will presumably drive away the better teachers.
This last concept also applies to the lesser universities. I would much prefer to attend a best-of-breed online course than the drivel that I attended years ago.
This all mostly applies to courses that aren't mandatorily hands on such as a chemistry class. Even if excellent simulators and whatnot were developed it would be very odd to have a chemistry degree without ever setting foot in a lab.
The last bit comes from acceptance. Many top universities have things such as SAT requirements and high-school grade requirements along with a somewhat mysterious set of criteria to get in. But where this could also get interesting is if a student is doing very well with a top universities on-line system, yet didn't really meet their other classical requirements could this potentially be another new route to entry?
But it will be at the high school level where the most real change will be happening the soonest; regardless of the wishes of the high school administrators.