Right. And that source isn't claiming that people are all equally good at founding startups. Rather, it's claiming that previous failures don't substantially make a person better at founding successful companies. (They do find a very small 2% effect, but that may be statistical noise).