During startup school Zuckerberg said something along the lines of 'a company is the most efficient way to align a large group of people to solve a problem' and

During startup school Zuckerberg said something along the lines of 'a company is the most efficient way to align a large group of people to solve a problem' and this is going to be the fundamental obstacle to a true crowdsource model. You need to be able to build up your domain knowledge as a company, not just tap into the knowledge of others temporarily. There is also the 800 lbs gorilla of support which can probably never be effectively solved via a pie-in-the-sky everyone does what they enjoy model. Pursuing your passion is simply not scalable, sure there will be plenty of people that get away with it but it is not feasible to say everyone will be able to. Someone needs to mop the floors, clean the toilets, and answer calls to the help desk. At the end of the day I agree that freelance will probably become more prevalent and the commoditization of software components will certainly reduce the need for in-house expertise. These forces will bring us closer to the authors vision but the demand for support and in-house knowledge will herd us into monolithic organizations for the foreseeable future.