Morphing The French

BACK on July 17 I began a seven-part series (https://beauchess.blogspot.com/2019/07/morphy-v-french.html) on how Paul Morphy took on the French Defence.
Since then I have played in a 4NCL congress in Telford, the British, a CzechTour festival in Olomouc, Tim Wall's Northumbria congress and two games in London's Summer League.
Only today, in round eight of the 'Paignton' premier, did I get a chance to put what I learnt into practice.
Spanton (1881/168) - Alan M Brown (1991/186)
French Exchange
1.e4 e6 2.d4 d5 3.exd5 exd5 4.Nf3 Bd6 5.Bd3 Nf6 6.0-0 0-0 7.Nc3 c6 8.Bg5!?
This move is not liked by the analysis engines Stockfish10 and Komodo10, but it was played by Morphy in wins against Johann Löwenthal and the Cuban slave José María Sicre. The engines like 8.h3 or 8.Ne2.
8...Bg4
The engines' choice. Löwenthal preferred 8...h6 9.Bh4 Bg4 in game 10 of their match.
9.h3 Be6
Morphy - Sicre, Havana 1862, saw 9...Bxf3 10.Qxf3 Nbd7 with a slight edge to White thanks to having the bishop-pair (1-0, 27 moves). The main move is 9...Bh5, when the double-edged 10.g4!? is interesting.
10.Ne5 Nbd7 11.Nxd7
11.f4 looked dangerous in the postmortem, but we did not consider the engines' suggestion of 11...Qb6!? 12.Kh1 Qxb2, although White has lots of compensation for the pawn.
11...Bxd7 12.Qf3 Be7 13.Rfe1 Be6?!
More natural is 13...Re8.
14.Ne2 h6 15.Bh4 Nh7 16.Bxe7 Qxe7 17.Bxh7+?!
Keeping more material on the board with 17.Nf4 looks a better try, but the text is enough for an edge, but one that turns out to be largely meaningless.
17...Kxh7 18.Nf4 Qf6 19.Nxe6 fxe6 20.Qxf6 Rxf6 21.Re5 Raf8 22.f3 Kg8 23.Rae1
White is slightly better because Black has a backward e pawn, but the reduced material means the weakness is easily defended without having to make concessions elsewhere (½–½, 30 moves).