A Game and a Puzzle, Revisited
(Position after 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ Kxf7 5.Nxe5+ Nxe5 6.Qh5+ g6 7.Qxe5 d6 8.Qxh8 Qh4 9.O-O Nf6 10.Qd8)
Rick Kennedy has offered a very useful comment, helpfully providing notes from several other sources. In reply, I'd first like to express my gratitude to Mr. Kennedy for taking the time to write, and I look forward to investigating the links he has included.
As for the analysis, most of it repeats what I presented, but three lines are worth mentioning.
(1) 7...Qe7! I hadn't explicitly said it, but I certainly implied that Blackburne's 6...g6 was objectively a mistake, played for the sake of "coffeehousing" his opponent when the simple 6...Kf8 7.Qxe5 d6 leaves White without any real compensation for the material. I still think 6...Kf8 is a good move, but I didn't even consider 7...Qe7 (after 6...g6 7.Qxe5). If White doesn't take the rook, then the situation is essentially the same as it is after 6...Kf8, while if White does take the rook, it's pretty easy to prove that Black wins.
(2) In my main line, Kennedy, citing analysis by Geoff Chandler and Todor Dimitrov, varies from my 12.Qxb7 with 12.gxh3, showing that it likewise draws after 12...Qxh3 13.Qxb7 Qg4+ 14.Kh1 Qf3+ etc. or 13...Ng4 14.Qxa8+ etc. (Note that Black can't escape the checks with 14...Ke7 15.Qb7+! Kf6?? [15...Kd8/e8/f8=] because of 16.e5+ followed by 17.Qg2.)
(3) Chandler & Dimitrov also mention 12.Qxb7 and suggest it loses, but the culprit is not 12.Qxb7 but their 14.e5?, after which Black has a forced mate.
Very interesting and I'm grateful to Kennedy for his comment...but my dream remains unfulfilled - can't Black win after 10.Qd8, somehow?