May be the case - but in my experience, issues of scale don't creep up on you - one day hit a hard limit and your service is crashing all the time.
.. And if you haven't thought about it - this can happen at any volume. I've seen apps that test great with 5 users and then bomb when you get 3 more.
Just my experience. Not saying as an overall strategy it's not a good idea - as a business this might very well be an effective model - just saying that ignoring it technically is a bad one.
Interesting - he makes a point about not giving too much away for free (literally at the very end of the talk).
He mentions that after putting their book up for free that they didn't sell many books any more (although, arguably the pay-off was exposure, conferences, etc...).
For example "Give up on hard problems" "work less" "look for easy pickings at the low end of big companies" etc.