unless I'm misunderstanding the argument, I dont see how those hypothetical returns could be considered "expected returns". startups which reach a place where employees can profitably cash out seem far too rare to reasonably expect a return at all, never mind a large one.
Since a person works for one company at a time (usually), and it can take 3-5 years or more for a startup to reach a place where the equity is worth something, this argument reads to me like "the returns on a Powerball win are so much higher than your projected lifetime earnings that playing the lottery is a smart financial move".