I just watched this
The talk is short and very sweet, with a nice amount of humour. I had one of those 'Yes....exactly!!' moments.
I've often tried to use programming and computational approaches to compensate for gaps in my statistical foundations. I was exactly like so many of the folks he referred to. Yes, I passed various levels of Stats course through my Engineering college track, and passed well in some cases (!), but there were many residual weaknesses. It wasn't till I started working and trying to crunch data to gain understanding, that I started to 'get it' in any meaningful way. That learning journey is never ending.
Simulating processes was one way I successfully gained understanding about processes. Nowdays, there are all manner of great tools to help, from high-power simulation packages, to bare-bones programming environments with emphasis on statistics.
The combo of great programming chops and such tools lowers the barriers to entry, for those not so inclined to be bothered with the deeper statistical theories, or maybe even with simply, better things to do!