Interesting read, but I have one major qualm about this. There seems to be a confirmatory bias among the academic community that a measure of intellectual success can be found in the percentage of individuals who earn a PhD. Sure, it's true that you need to be reasonably smart to earn a PhD, but I think that someone can have a lot of intellectual success (potentially just as much or maybe more) if they don't earn a P…
Like say IQ. Feynman wouldn't have qualified for Mensa, and happily too[0]. Maybe that's a good metric. The metric is happy not to be in Mensa, but also has a PhD.
Edit: Feynman's IQ was 126 and Mensa wants 132, at least according to Ericsson and Pool in their book Peak.