I have some very strong doubts that a MOOC will ever be able to offer the laboratory experience that you'd get at a traditional university (I'm not talking entry level). To do a lot of the EE work I've done in the past, you really need to know what you're doing, and a lot of the equipment, like a good vector network analyzer, is both incredibly cost prohibitive (new ones can reach a quarter of a million), and really take some expert knowledge to operate effectively (i.e. someone should be there with you to ensure the you don't break the machine). MOOCs may be good for CS, and they may be good for teaching entry level knowledge, because of the way the accumulation of very expensive equipment and experience/intelligence surrounding the field, at least for the immediate engineering fields I've been exposed to. I think it's still far better (if not the only decent way) to have a faculty mentor present and have a lot more close cooperation (as atleast I did when I was in university) when learning a lot of the specialist EE knowledge I got back when I was in school. I doubt you'll ever see MOOCs progress much in this regard, nor would I want to hire an engineer who did not have some level of hands on design and lab experience described above.

Also to deflect a potential counterpoint: most of the very expensive equipment will probably never be any cheaper because it isn't built in enough quantities, and for things like good RF equipment, everything needs to be carefully spec'd and machined, and the manufacturing is very time and resource intensive (you're getting your money's worth on a good spectrum analyzer or network analyzer, typically).