Interesting article on NYTimes on the subject:
1) The problem with most universities around the world is that most teachers do no add real value to the education process, they just repeat content in a way similar to a virtual course. It is not uncommon that the content is obsolete and ignores world class best practices.
2) Facebook, LinkedIn & Co. have already proved that networking can be initiated virtually. Of course face to face relationships are necessary to develop trust and confidence. But virtual networking is an excellent starting point and travel has become economically available.
3) Talent is spread out through the world, and most companies have difficulties identifying the talent. Talent shortage is a common problem, not just in G8 countries but all around our planet.
I think the greatest challenge of MOOCs is to produce world class content, this requires some "superstars" but in most cases just very good teachers. In many fields content quickly becomes outdated. And talented people that also are productive regularly update their knowledge.
Identifying this very good teachers is probably the greatest challenge of MOOCs and government education agencies. Teachers have "titles", categories, "names", artificial reputations, etc. that can make it very difficult to identify the ones that really excel in teaching.
Identifying talent and productive people is a great business for MOOCs. They do not compete with Harvard & Co., they reach out to a talent source that due to economical and relationship constraints they can never tap into.
MOOCs do not compete with top universities, the focus on another segment in the business of identifying talent. They will be successful only if they provide top instruction at Harvard level.
Open Source does not focus on the economical elite, it is open to whoever has the talent, will and discipline of taking advantage of it. Now imagine this last statement on developing professionals that have the talent, will and discipline to take advantage of MOOCs.