I honestly didn't think it worked out and most people I talked to after the session didn't think it worked either. To put it lightly: People had rather mixed impressions.
He was progressively writing Node.js-code without explaining any of it in detail. His way of illustrating what his code was doing, was by cross-referencing his Node-stuff to some Ruby-based BDD framework.
How do I connect that to Node? Do I need Ruby to run node?
It came off wrong to most people. Especially with the techo-music glaring in the background, some people thought he just wanted to show off how cool he was and brush his own ego.
Other's complained that if you could actually follow his demos, you already knew the stuff he was demoing, so what was the point? So what had you just learned?
And like he said: Some people directly told him it was a waste of time.
IMO it was an interesting idea taken too far.
He makes it come off as a success, but I think that is disingenuous. This session probably disappointed more spectators than any other session on that conference. And I say that as someone who knows how good Rob can be when he's at his best.
So don't be afraid to experiment. But be prepared to see your experiments fail. Every now and then they will. Even for a guy like Rob. Especially when you take it to the extreme, like he did here.