As given by LeClown, 8...Ne7 transpose to main line and game Wojtaeszek-Halkias,Capelle la Grande 2002 (I played one simul game as Black: with 9.Qb1!? Nbc6 10.Nf3 a6 - Iordachescu-Laurain,Naujac 2002).
But it seems White always has a small advantage (I think the variation is a better version of Kasparov's gambit since 9.h5 h6 10.Nf3 Nbc6 11.Rh4! protect the 4-rank and the d4 pawn; I recall Kasparov's gambit: 6...Ne7 7.h4 Qa5 8.Bd2 cxd4 9.cxd4 Qa4 10.h5 Nbc6 11.Nf3) thus I will prefer 8..Nc6 9.Nf3 (9.Qg4 g6 with attack on d4 and c2) cxd4!?

It seems that 8.h4!? is a novelty. Before, white almost always played 8.Qb1, 8.Nf3 or 8.Qg4. 
Playing h4 in the opening seems to be the recent trend to meet the french. For example, a similar idea occur in the "Shirov/Anand" line (3.Nc3 Nf6 4.e5 Nd7 5.Nce2).
In my opinion, Shipov's 8...f6 is a mistake, or at least a very risky move. There was no need to undermine white's center while black is undeveloped. Moves like 8...Ne7 or 8...Nc6 suggests itselfs, completing the development before taking measures against white's center. After 8...Ne7, the game would have return in known territory. In fact, this position is often reach via another move order: 3.Nc3 Bb4 4.e5 c5 5.a3 Bxc3 6.bxc3 Ne7 7.h4 Qa5 8.Bd2 Qa4...
So, we can say that Shipov has been "move-ordered" by Morozevitch.