I think they doubled their leverage before their genius plan went off the rails. They were over leveraged and hugely over confident and not even watching the business very close. It's fun to watch when academics step out of their universe where they can spend a lifetime being wrong and still get their retirment.
It does NOT matter he is MIT or phsycist. If he can make annual 20% for sure, then 1) he should show 100*1.2^5 = 248.832 for the last 5 years with initial seed as much as 100. 1) he should show 100*1.2^10 = 619.1736 for the last 10 years with initial seed as much as 100. He can continue for the next 50 years to make 100*1.2^50 = 910043.8 which is 9100 times than initial seed, like Buffet's record.
"How can this be profitable selling at the money straddles in addition to buying wings on those straddles and using stop loss pending orders of the underlying?" I don't think he buys the underlying so a stop loss order can't be put on it; the stop losses would be on the options. I've heard that selling iron butterflies is a proven profitable strategy either with or without stops.