If you agree with point 27 (questing for Minimum Viable Product/Experience), then rule 14 makes a lot of sense to me. He isn't saying not to design a full web interface (which as you point out, can do more). He is merely saying by doing mobile dev first, you are essentially forced into discovering your MVP more lucidly. You can then add features to the full web interface from there.

>A full web interface can always do more and the challenge should be converting those features to mobile as opposed to dumbing down the design for the lowest denominator.

I think that this sentence gets to the crux of his point. Terms like 'dumbing down' and 'lowest common denominator' to describe the mobile experience seem a little loaded to me, and imho, this is exactly the type of situation where deving movile first could provide surprising insight into your MVP. I'm talking honest to goodness slap in the face reveals.

I think it's safe to say we have different opinions on this point, but man...I simply can't count the number of "Hey check out my startup" websites that I've clicked on and just kind of had my jaw drop at what was going on. I'd honestly say the majority of the ones I've seen could have benefited from this approach. That's completely anecdotal, I know, but it's the impression I have.