@gewgaw: In your analysis White can apparently vary his play and prefer a somewhat safer set-up:
1. e4 e5 2. Nc3 Nc6 3. d4 Nxd4 4. f4 d5 5. exd5 Qe7 6. Bd3 exf4+ 7. Be4 Qh4+ 8.Kf1 Bc5 9. Na4 Bg4 10. Qd2 Be2+ 11. Nxe2 Nf3 12. Nxc5 Nxd2+ 13. Bxd2 Qe7 14. Bb4 Nf6 15. Bf3 0-0-0 16. d6 cxd6 17. Nxb7 Rd7 18. Na5 Qe3. Here you continue with 19. Rd1, but
19.Nc6!? followed by an eventual Nd4 looks solid and good: White keeps his pieces together and defends easily. - However, there is still 9...Qe7 when White has more problems to overcome.
@Markovich: I haven't checked your
7...Nf5 8.Bxf4 in detail. The reason why I like
8.Nge2 is that it invites 8...g5, similar to a King's Gambit. White appears to be OK. If Black avoids to protect pawn f4, White can still take it, maybe under better circumstances.
My Fyfe research began a few years ago, this story seemed worth to be told, no matter whether 3.d4 was entirely correct. I found sources and old games. Markovich's idea 4....d5 shed new light on the topic, which was highly welcome. However, in my view 4...d5 leads to an open battle, similar to other classical gambits which are still unsolved after 150 years or more. This part of the Fyfe is full of sharp tactics, for both sides.
Severino wrote on 07/14/12 at 10:12:22:
2. After 3.d4 Nxd4, it is White, rather than Black, who has his work cut out to get an even position. This could certainly be an argument against playing the Fyfe Gambit, though the positions seem to be complex enough to give chances in practical play.
Yes, it is an experiment. The computer brought new insights: piece activity and open files are worth a lot. But 3.d4 is still an extreme case.
Severino wrote on 07/14/12 at 10:12:22:
6. Considering the number of controversial openings that have had their advocates, it seems hard to believe that, following Peter Fyfe, the Fyfe gambit seems had no champion, if only for a brief period of time. Openings are played for a variety of reasons. Has there subsequently been no Glasgow player who has taken up the Fyfe Gambit? Perhaps there was, as Stefan Bücker mentions a later thematic tournament.
Perhaps the 1917 tournament was sponsored by Fyfe, who died in 1940. No foreign participants, as far as I know, which isn't surprising during a war. - The Fyfe gambit makes sense only when you trust the Belgrade Gambit. The Belgrade Gambit was played only in 1945 by Trajkovic (a first attempt in 1938 by Kurt Richter was a failure) and fully rehabilitated only decades later by Bruce Monson.
Thank you very much, Severino, for sharing your thoughts. The sources you mention (Miller, Pierce) are important. I had already mentioned Google Books, many sources are available. The British Library "delivered" two games from 1917.