Believe me or not, but when I played the Ryder Gambit some 15 years ago (only OTB) I found that there were a few points not entirely clear after 6...Qg4.
The line that bothered me most was 6...Qh4+ 7.g3 Qb4 (an improvement on 6...Qb4) 8.0-0-0 c6 and as White can't play 9.Qg3 Black is better.

Let's get a bit generous and offer two pawns.
I own a copy of Diemer's book "Vom ersten Zug auf Matt". It largely deals with a gambit idea Diemer credits to a Dr. Ryder.

1.d4 d5 2.e4 dxe4 3.Nc3 Nf6 4.f3 exf3 5.Qxf3

From theoretical point of view I can't believe that this is good anyhow. But Diemer played it quite regularly. And on practical basis it worked.
The question to me is: What ist the best way to defend Black's case in this variation? Does White get enough practical chances (not theoretical), or is there a clear refutation - discovered by, say, Rybka?
Can Black simply take ond d4 (5...Qxd4) and how does he proceed after 6.Be3? 
Are other continuations - 5...c6, 5...e6, 5...g6, 5...Nc6, 5...Nbd7 - more to the point?

cheese