His choice to pair 1.e4 e5 as Black leading to the Breyer with the Grünfeld also surprised me. Still after some thought it makes sense to me if you read some of his white recommendations for the 1.d4 repertoire.
Let's check a couple of alternatives:
1. Semi-Slav: He considerers the Semi-Slav slightly problematic because of 5.Bg5 plus move-order issues incl. Slav Exchange. I guess this also puts him off the regular Slav.
2. Nimzo/Queen's Indian/Catalan: he doesn't state this explicitly, but I guess it's hard to get this on 100 pages plus it is difficult to combine with the repertoire vs. 1.Nf3 and 1.c4. In addition it's hard to find something in the Catalan at all where you can claim black preserves some winning chances.
3. King's Indian: He quite frankly states that he thinks it is a not completely sound opening. Of course that's debatable.
4. Queen's Gambit Declined: I think this is just too 'unsexy' to present plus you have the Catalan issue and general lack of winning chances. Sure, you could argue here (I play this myself as black, so dont accuse me

).
In general he has the problem of showing a repertoire for Black and White, so there is bound to be a point where both sides meet, so to say. Here it is the Russian Variation of the Grünfeld, where he basically slighty prefers white but states that the position is sufficiently unbalanced to offer black chances as well. I think the Grünfeld is maybe the best opening to come up with such an overlap. For instance if he would have choosen the QGD Lasker or black side if the Catalan the end of the variation is either a stale complete draw or some pull for white, but never any 'imbalances'.