but that all seems redundant about opening the diagonals when you are opening the h-file to your own castle and yes black looks solid but white can build up an attack while blacks pieces are tied down to defending his king

The position after 12... Bg7 looks very comfortable for black. White has no attack and it is unclear where white will put his king. Instead, after 9...fxg6 black has to worry about tactics on d5, the e file opening with threats against e6, the a2-g8 diagonal (and a1-g7 diagonal), and still the h file, so white can claim an advantage (or at least more practical chances).

Houdini 1.5 x64 at depth 22 gives hxg6 as 0.46 white edge and fxg6 as 0.49 white edge.

However, the position after 10. Ne5 dxc4 is approximately equal, as Qf3 is met by Kg7.

(My analysis without the computer was flawed, after Re8 Qf3 Bf8 is met by Rh8+ (the only move that keeps any white advantage, not counting cxd5 exd5 Rh8+ with the same idea.)

I think Nh5 is a blunder.Here is my analysis:-

Also, Bf4 transposes to London System and not Bc4. 

I liked the commentary on the hxg6 v fxg6 today in Svidler Gelfand by Nigel Short: "Dragon players always play fxg6"

How true, how true.

i play the dragon !! what a coincidence

I have a friend rated 1700 who specializes at the dragon (he probably plays the dragon variation at about a 2300 level), and he always plays fxg6.

i cant guarantee at what level i play it 

im not sure if this is the reason why but i think it might be because they are not worried about opening the diagonals since the dragon bishop covers them also they do not want to open the H-file