This appears to be a sound analysis to me.
TL;DR if you a priori believe that there is cheating in chess--as can be measured by correlation to top move engine play--then Hans has a high probability of being one of the cheaters. (if you believe that there is not cheating, then this analysis confirms that)
