I think the author messed up a classic "Doomsday argument" [1] style estimate which wouldn't depend upon the distribution. I see no way to recover the attempt,

I think the author messed up a classic "Doomsday argument" [1] style estimate which wouldn't depend upon the distribution. I see no way to recover the attempt, though. (I assume that's what the author was going for, given the precision of the estimate along with the use of 95% and 20% which are almost the standard doomsday argument values: that we're 95% certain that the total number of humans ever is at most 20 times (not 20%) the number that have existed before us.) [1] https://en.wikipedia.org/wiki/Doomsday_argument https://en.wikipedia.org/wiki/Doomsday_argument