I've seen this talk before, and he makes a ton of assertions mostly backed by often-repeated cliches rather than data (e.g. participation trophies). This is a really good talk I've seen posted in response to this very video before, so I'll post it again.