I would argue that three 22 year olds on their first startup shouldn't be building a product that can't be validated by the market in one or two years.

Define validate. That is kind of the crux of my question. If it means "profitability" then that is a different threshold than "users." I am trying to figure out what PG is trying to describe.

It's pretty obvious when it happens but it's most definitely not profitability. Rapidly growing usage (ie, traction) with an envisionable business model is usually sufficient. While the revenues themselves are not strictly necessary, it does help to demonstrate the ability to collect them.

I an not too sure about this. The debatable issue is should any investor give them any money or not? Since I doubt it almost never happens we probably don't have the data to know if it would work or not.