It's always hard to figure out what exactly to look at... I find both things plausible, both that the most successful startups are not started by the "rich kid

It's always hard to figure out what exactly to look at... I find both things plausible, both that the most successful startups are not started by the "rich kids" and that the "rich kids" start most successful startups. The paper you linked is talking only about the first thing, by limiting its analysis to startups that surpass a $1B valuation. But you could choose some much more modest definition of "successful startup", like "all startups that reach series C or have an exit that returns at least 1.5x", and see totally different results.