Without a doubt - one of the goals should be to reduce the total cost of a college education by at least a third if not more - and to do this by the end of this decade if not earlier.
Some have implicitly argued that the problem is not the total cost but the productivity of the education sector. By this argument, as long as average productivity rises (perhaps even as tuition continues to rise) the education sector is not doing too badly. The strategy therefore is just to have as many people watch their videos as possible.
Average Productivity = Number of people watching videos / time needed to make videos. With a fixed denominator, average productivity will naturally rise.
This argument is detached from the usual widget argument which also claims that as total output expands costs per unit drops. In a perfect competition, prices also fall. Certainly the education sector is as far from perfect competition as I can imagine but any claim that MOOCs increase productivity must also be accompanied by a fall in tuition.
The clearest way to decrease college attendance costs is to substitute expensive college courses on campus with free MOOC courses. Next comes the details - how to transfer the credits. The MOOCs have honed in on two: accreditation and certification. They are different sides of the same coin to me but essentially the idea would be to make the student pay in order to receive a piece of paper (electronic or otherwise) that allows credits to be transferred. This would be accompanied by (most likely) a proctored exam that the student has to take in order to prove competency.
Of course this assumes that college education as we know it now continues to have the same relevance not just in the workplace but in our lives. By this I mean we go to school, graduate, get a job, and forget almost all about what was taught in the classroom. Somehow, the ideas that MOOCs have spawned should require them to reach higher than that.
Unfortunately, it seems that they may have been captured by the vultures of Sand Hill Road and need to monetize their experiment as quickly as possible. If this is what is to be then the goal of reducing total cost of attendance could even be considered redeeming.