Was indeed afraid that there wasnt one book. Too bad as it will mean much more work

The thing is that with 1 c4 or Nf3 (or g3) move-orders, there are many move order tricks to possibly avoid a Grünfeld, which I prefer over the fianchetto variation of the KID. 
Like in Glenn/mnb's D variation 1.c4 Nf6 2.Nc3 g6 3.g3 Bg7 4.Bg2 0-0 5.Nf3 d6 White can now play 5 d4.
There are many of that kind of tricks where you can avoid the Grünfeld and certain variations of the KID as white.

The King's Indian player who answers 1.c4 with 1...g6 isn't always looking to transpose to 1.c4 e5 positions (only when it suits him).  Many English players use 1.c4 as a tranpositional device to avoid certain 1.d4 openings (for example the Gruenfeld).  The King's Indian often isn't avoided and black gets what he wants.

It's not really a question of advantages or disadvantages between 1...Nf6 and 1...g6 vs The English.  

The real difference between the two moves, at least for KID players, is one of flexibility. As a Kings Indian player myself I consider 1...g6 a more flexible way to meet the English, since in some lines its quite nice to have the option to go Nge7. Having said that, I think its largely a matter of taste, since KID players have little to fear in going 1...Nf6 either.

Another reason why I prefer 1...g6 over 1...Nf6 is that Kings Indian players tend to have very aggressive styles, and 1...g6 generally gives better chances of a more complicated, red blooded battle against the amorphous and stodgy English Opening. 

One of my favorite KID players GM John Nunn, also prefers the 1...g6 move order so as to be able to get reasonable winning chances against the following annoying and ultra solid setup featuring e3, Nge2, d3  etc. Here is a sample game from his praxis:  

[Event "OHRA"]
[Site "Amsterdam"]
[Date "1985.07.??"]
[Round "10"]
[White "Sunye Neto,Jaime"]
[Black "Nunn,John DM"]
[Result "0-1"]
[Eco "A25"]
1.c4 g6 2.Nc3 Bg7 3.g3 e5 4.Bg2 d6 5.e3 Nc6 6.Nge2 h5 7.h4 Bg4 8.d3 Nge7 9.b4 a6 10.Rb1 Rb8 11.Qc2 0-0 12.Bd2 Qd7 13.Nd5 b5 14.a4 e4 15.Nxe7+ Qxe7 16.Bxe4 Ne5 17.cxb5 axb5 18.axb5 d5 19.Bg2 Nf3+ 20.Kf1 Rxb5 21.Ng1 Nxd2+ 22.Qxd2 Rfb8 23.d4 c5 24.dxc5 Qxc5 25.Bf3 Rxb4 26.Rxb4 Rxb4 27.Qxd5 Rb1+ 28.Kg2 Qc1 29.Kh2 Qf1 30.Bg2 Qxf2 31.Nh3 Rxh1+ 32.Kxh1 Bxh3 33.Bxh3 Qxg3 0-1

Quite an inspirational and important game in this line, and a good way to challenge White's super solid setup. Those of you who want a little more insight into this line  should get "John Nunn's Best Games" by the man himself, the annotations are deep and enlightening.

Adios amigos

Top   

Glenn Snow doubts the statement, that White can take benefit from the move order 1.c4 Nf6 2.Nc3 g6 3.g3 Bg7 4.Bg2 o-o. As I am certainly not an expert, I have done some database statistics. As this has some objections, I must note that all results are based on 50 games at least and usually on several hundreds.

A1) 5.d3 d6 6.e4 e5 7.Nge2 Nc6 8.o-o 44% wins, 31% draws and 25% losses; score 60%.
This is probably the reason, that the move order 1...Nf6 is disapproved by theory.
A2) 5.d3 d6 6.e4 c5 7.Nge2 Nc6 8.o-o  32%, 32%, 36%; score 48%.
This confirms Glenn's opinion.

B1) 5.d3 d6 6.e3 e5 7.Nge2 Nc6 8.o-o  36%, 31%, 33%; score 52%.
B2) 5.d3 d6 6.e3 c5 7.Nge2 Nc6 8.o-o  37%, 33%, 30%; score 54%.
This should not scare Black.

C1) 5.d3 d6 6.Nf3 e5 7.o-o Nc6 8.Rb1  34%, 36%, 30%; score 52%.
     But after a5 9.a3 Re8 it is even 23%, 45%, 32%; score 46%.
C2) 5.d3 d6 6.Nf3 c5 7.o-o Nc6 8.a3  35%, 44%, 21%; score 57%.
This amazes me, as C2 seems a boring symmetrical system to me.

D1) 5.Nf3 d6 6.o-o e5 7.b4!?  41%, 35%, 24%; score 59%.
D2) 5.Nf3 d6 6.o-o c5 7.d4  44%, 33%, 23%; score 60%.
D3) 5.Nf3 d6 6.o-o Nc6 7.d4 is a transposition to a line of the Fianchetto Variation of the KI, which Black might not have on his repertoire.
D4) 5.Nf3 d6 6.o-o Nbd7 7.Rb1  51%, 33%, 16%; score 68%.


So with some hesitation I stick to my statement, that 1...Nf6/2...g6 is less precise than 1...g6; not because of the Botvinnik System, but because of 5.Nf3. After 1.c4 g6 2.Nc3 Bg7 3.g3 d6 4.Bg2 e5 5.Nf3 Nc6 6.o-o Black has f5 or Nge7.

You will probably have to look in books on the English and Réti. Two remarks: after 1.c4 for a KI player g6 is more precise than Nf6, in case White plays the Botvinnik System. If White does not play d2-d4, Black will play e7-e5 transposing to a Closed English (compare Closed Sicilian) or c7-c5, transposing to the Symmetrical Variation.
Concrete: 1.c4 g6 2.Nc3 Bg7 3.g3 (3.e4 d6 too) d6 4.Bg2 e5 and Black has kept all options open.

I'd recommend picking up acopy of Gallagher's Beating the Anti King's Indians to cover the out of the ordinary stuff like the London, Barry, Torre, Tromp, Veresov and BDG. It is excellent (typical Gallagher) and while you are at it his Stating Out: The King's Indianis very worthwhile. Personally, I buy almost everything I can get my hands on dealing with the KID, but I would recommend those two and this one over all of the others.

Thanks Ed Gaillard, X and Top Notch!

I read the whole book and at least 80% of the variations, and it's exceeded expectations  Simply fantastic. Of all opening authors, he has to be among the top 3, if not first. 

It's true he doesn't cover the London attack, or Veresov etc. That's too bad, but space is limited, I guess. However, I found a way around that...I looked up Gallagher's games in a database and saw what he plays against those systems!  8)

Great book!