this has already been discussed on the sub a couple of times. essentially the data points seem to be pretty accurate but there are interpretation problems in the conclusions he draws and the nature of the comparison is apples to kumquats not apples to apple