The book I read said 40, so that's my "first and best" truth. All others require "unlearning" and are therefore suspicious.

Revising statistics is pretty common - it's one of those cases of the observer influencing the observation. Like happily ever after endings, unbiased data colletion is a myth we hold too dear. We make our data dance to the tune we choose. Very arbitrary.

The other number I heard was "half a billion people infected." - so some recovered and some never appeared sick.