NVM that suggestion. I read the moves wrong. The knight is gone. I fixed the 2 analyses above.
Traxler QE7 Lines virtually ALWAYS WIN in KF1 Variations

There are so many attacking lines
15. Ng1+ 16. Ke1. Now what?
17. Qxg2 Just play Re1. He can take the h pawn if he wants it. White still has three minor pieces and a rook for the queen, which is plenty.
Black has 3 passed pawns on kingside, whites queenside is all blocked in, white has to worry about mating nets with the knight and queen, as well as the a8 rook at some point.

It's hard to tell from this position, black has all those pawns, but black king is in some danger. Black should start storming those pawns down the board before white fully develops his pieces and try to break through in time.

But how about 9. d6 instead of 9. c3?
Pawn just takes, and the threat of bg4 and qf6 is still inevitable and black king can escape checks by the bishop
Pawn just takes is just a mistake. Taking with the queen is correct, and Black is not worse if he knows his homework.
But who cares about the stupid 5.Nxf7?
Black can easily equalize and white is behind in development, AND black has potential tactics
Ι'm sorry to say that your analysis & evaluation are horrible. Maybe you wanted to say "white has attacking possibilities, as well as an extra bishop"?
The key to hold the other position (after 5.Nxf7?) actually is Black's 20th move here - 20...Qg6!
Tested in several correspondence games, and Black is doing OK. Actually 10...Qe7!? (in place of 10...Qc5) does look like a cleaner way to equalize.
And- last, but not least- I believe that white's strongest move is 5.d4! with a clear advantage (insert some 12-15 moves of well-established theory).

I want to see how the computer plays out my analysis after qxh2. I'm on touchscreen device now so I can't input positions.

Queen and 3 pawns? I'm afraid you have to count again... (#24): both have queens, white is temporarily a pawn up (very soon a piece) and he/she does not have to take the c6 knight at once. Why not 10.d3 first? Black will lose a piece for no pawns, and his king is more exposed to something like Rh1-f1+ than white's (who's safe at e1). Pieces being equal, white would have the advantage. With a piece in pocket for white, this is not interesting at all...
If you mean #34, well, this is even less interesting: rook and 3 pieces for queen and (sometimes) one pawn. Yawn...

And- last, but not least- I believe that white's strongest move is 5.d4! with a clear advantage (insert some 12-15 moves of well-established theory).
Did you find a concrete continuation for White in the 5...d5 6 Bxd5 Nxd5 7 dxc5 Ndb4 8 a3 Qxd1+ 9 Kxd1 Na6 10 b4 Nd4 11 Be3 b6 line? Otherwise Black seems to have reasonable play here.

And- last, but not least- I believe that white's strongest move is 5.d4! with a clear advantage (insert some 12-15 moves of well-established theory).
Did you find a concrete continuation for White in the 5...d5 6 Bxd5 Nxd5 7 dxc5 Ndb4 8 a3 Qxd1+ 9 Kxd1 Na6 10 b4 Nd4 11 Be3 b6 line? Otherwise Black seems to have reasonable play here.
Might depend on your definition of "reasonable". As far as I'm concerned Black is a pawn down for very sketchy compensation. A Ruy Lopez player would sell his house to get such a position against, say, the Berlin Wall.

IM pfren, you're right, I played out that d3 live and white won on stockfish. As far as that BH4 line, what are all the computer moves after?
I just played a game exactly like this and got 100% accuracy.
https://www.chess.com/analysis/game/live/71925216821?tab=review
There are so many attacking lines
15. Ng1+ 16. Ke1. Now what?
17. Qxg2 Just play Re1. He can take the h pawn if he wants it. White still has three minor pieces and a rook for the queen, which is plenty.