The flaw in Ellsberg's reasoning, termed "selection bias" in statistical circles, is that he assumes that the startup founders he named are similar to their peers who chose to stay in college in all categories except their choice to leave college, and therefore, that leaving college itself was the cause for their success. Rather, these highly talented individuals likely differed from their peers on the exact characteristics Ellsberg thinks education erodes -- networking capacity, marketing, comfort with failure, and of course, innate creative talent -- well before they ever chose to leave college.
His point is all the more unfounded when you consider each of these unique people's cases individually: It's possible that Mark Zuckerberg would never have been successful if he hadn't gone to college -- it was only at Harvard that he famously recognized the niche which Facebook now occupies. And in his 2005 commencement address at Stanford University, Steve Jobs credited his unique penchant for design to a calligraphy class he had taken at Reed College.
