Yes, there is a bit of a self-selecting bias involved here, but the thesis of _WHY_ you should take less valuation is (in the linked article) not based at all on the notion that you then end up with a large piece of the success pie.
It's based on: Hey, take what you need to prove that you have a good product, and don't take more, because taking more just means you'll be chasing a failure for that much longer.
The central tenet of "stop worrying about runway so much, you want as much runway as you need to prove your product and any more runway is just wasting everybody's time. Runway is a tool; not a goal." sounds rather compelling to me, at any rate.