Actually I do not think this is a well played game, by either side (my opponent is an ICCF IM).
I will probably not repeat 9...Qd7 again. While I would be very happy to see the established line (where white stands badly) my opponent played the very bold and risky, but positionally very consistent novelty 13.bxc3! after which I feel white should be "almost" OK.
15...Nfxh4! is a brave choice, which I had to play if I wanted to win. The other option was 15...Qe6, but after 16.h5 Qxa2 17.Qb5 Nh8 (17...c6 does not change anything) 18.Qb2! Black has a knight more for just a pawn, but white has terrific positional compensation- actually I don't believe he is worse at all.
16.Rdg1?! can't be a good move- both 16.Rde1 and 16.g3 were better moves, when Black should not have more than a slight edge.
19.Kd1? is the last straw. White had some hopes to be saved after 19.Be3. The rest of the game is a tactical mop-up, with the result never being in any doubt.
The above analysis is quite brief and inconclusive, I may expand and improve it later. The Halloween might be a crap opening, but refuting it completely still needs some work to be done.