I like his methodology, mainly because of a couple reasons. 1) We don't have that many civilization level societies to study in history, and 2) no society specifically sets out to implode, in fact there is tremendous benefit in not doing so. So even if we only had one collapse to study, we probably should do so and learn as much as we can from it, careful not to put too much value in the 'this time it's different' thing for our own prospects.