I'm certainly not saying 11. f3 isn't good. It's a powerful move that also has some drawbacks. I personally prefer to prepare this advance a little more with Rfe1.
I've found from my practical OTB experiences, it's very difficult for black to get play against white's e pawn. Maybe bad play on black's part, but I'm inclined to think if this variation was so strong, it would be more popular.
No, the e pawn won't be much of a target, unless white puts all of his pieces on the queenside! I think this takes away some of the flexibility of white's position, even though the 9 Nge2 line is praised for its flexibility. Black's real plan in my opinion would be like I said to breakthrough into white's weakened position with ...c5, the pawn weakness on e3 would only be an extra nuisance for white that makes his position less convenient.
I guess Nigel doesn't hate the fianchetto, not that much! I think immediately going for ...b6 has tactical drawbacks to a quick e4 but later on Short did put the bishop on b7, after making perhaps more desirable and less commital moves.
Despite white's excellent scores, I don't think this line is so clearly better for white (if it was, there would be less QGD. It is played less these days, but I don't think it's specifically because of the exchange, probably that white has many reasonable ways to keep black a bit passive and try for an advantage). Black's plans are harder to grasp, and I think he needs to play good prophylactic chess.
I still like the idea of putting the knight on e6, bringing the knight to h5 to remove the pin, and meet e4 with not only ...c5 maybe but also by putting a knight on f4. This way many things are still pointed toward white's slightly weakened kingside as well.
I think this line has been tried and tested at the highest levels