It is very waterfall-y as someone I know said in a talk and therefore you probably need to be a little careful what "rules" you take away from it. That said, the "mythical man month" discussion itself as well as the idea of a 9x difference in effort between a program and a programming system product are worth the price of admission.

The latter arguably runs somewhat counter to a lot of MVP, etc. approaches but recognizing the difference is still useful.