A while ago I wrote a blog article over at Chess.com about the Blackmar-Diemer Gambit, Ziegler Defence, and the related O'Kelly Defence.
Stefan Bücker and Lev Gutman have already established that White is doing OK in the line 1.d4 d5 2.e4 dxe4 3.Nc3 Nf6 4.f3 exf3 5.Nxf3 c6 6.Bc4 Bf5 7.Bg5 e6 8.Nh4!, for example see
Stefan Bücker's article http://home.lyse.net/chessmaster/StefanBucker_OKelly.pdf
But what does White do against 7...Nbd7, intending to develop with ...Be7 and ...0-0? Now 8.Nh4 doesn't work as well because of 8...Bg4, so White continues 8.Qe2 e6 9.0-0-0 Be7 (9...Bb4?! is met by 10.d5).
Now Guido De Bouver's blog (well worth a look, btw, for fans of the Blackmar-Diemer) gives 10.Bxf6 Bxf6 11.d5 in his article, "Not yet equality, but close", but concedes that Black might be slightly better after 11...Qe7.
At move 10 I prefer 10.Rhf1, putting the rook straight onto the f-file, a line which I recall discussing with "ArKheiN" at the Chesspublishing.com forum (via the move-order 7.Bg5 e6 8.Qe2 Be7, but unfortunately Black has the strong 8...Bb4! in that move-order).
I'm not 100% sure that White's compensation is objectively sufficient, but I would certainly fancy my chances with White in that final position. The g and h-pawns come rolling forward in the direction of Black's king.