Ziegler innovation

In my book "Attack with the Blackmar Diemer", I included some notes on an interesting line that was originally suggested by Stefan Buecker on chesscafe.com. ( In fact I called this the O'Kelly line, partly in honour of Belgian player Alberic O'Kelly De Galway who was the third correspondence chess worldchampion between 1958 and 1962 ). After 1.d4 d5 2.e4 dxe4 3.Nc3 Nc6 4.f3 c6 5.Bc4 exf3 6.Nxf3 Bf5, it was suggested to play 7.g4 !!

Taking the pawn with the bishop yields an even game, eg 7...Bxg4 8.Bxg7+ Kxg7 9.Ne5+ followed by 10...Nxg4

The sacrifice can be declined by 7...Bg6, aftre which white has the luxury of the choice after 8.Nh4, 8.Ne5 or 8.g5. As of yet, it is unclear what is best.

Accepting the sacrifice with 7...Nxg4 is obviouly critical, followed by 8.Nh4. This page will focus on defending the bishop with 8...g6
My book suggested 9.Bg5 but this seems to be insufficient after 9...h6
The complicated 9.Rg1 is suggested by Houdini

a/ 9...h5 10.Nxf5 gxf5 11.h3 Nf6 12.Qd3 and black is better (=+)
b/ 9...Bf6 10.Nxf5 gxf5 11.Be3 and black has no weaknesses (=+)
c/ 9...b5 10.Bb3 a5 11.a3 (=+)
d/ 9...Bh6 10.Rxg4 (=)
e/ 9...Qd7 10.Nxf5 gxf5 11.h3 with some white chances (=)
f/ 9...Nd7 10.Rxg4 (=)

So as far I can see, there seems to be no suitable white continuation after 8...g6 9.Rg1. Maybe Rg1 is wrong, but there dont seem to be much alternatives to complicate the game. So let's burry 7.g4 quickly !