I am surprised that, in the peer reviewed article, the authors do not defend the use of variance (their term - volatility) and the Sharpe ratio that require Berkshire's returns to be normally distributed for these statistics to be valid. I would not be surprised if a simple graph of the data would be very eye-opening, especially if compared with the comparable market data.

I would also expect the key result of this paper to be the death of CAPM and beta. To so blandly use a factor, Betting against Beta, as if the academic arguments Buffett and Munger have lambasted for decades were unreasonable, now that is news. Of course they both of them recognized this before Markowitz' original paper! Perhaps they should receive a Nobel Prize :)