Eleven questions about an englo scandinavian with opposite side castling

I played a rather long (51 moves, probably not so long at higher level, but mine generally end up much earlier, as I strive for tactics and imbalance) correspondence game where I missed a lot of stuff, and it proved to be hard to analysed, so chances are I still miss stuffs in my analysis.

Since the game (and my analysis) is rather long, I don't want you to review it fully (but you still can), but wil rather ask some more specific question.

Here is the analysis of the game :

An englo scandinavian (Malvinas). I think the opening was rather tense, and I think I miss a crushing line, then my opponet missed a very good line, then we went for opposite side castling with mating threats from both sides, my defense was precarious but sufficient, I had a bishop pair and more active pieces. I think I missed two opportunity to crush my opponent, and went for an endgame with only a small advantage for me, that I nearly blundered away. Then my opponent put up a strong resistance before blundering the game away.

The game was not played on chess.com, and the analysis there can have fancy arrows, so you might prefer go there :

https://lichess.org/study/SuNGs5RO/z5qj7vxo

Ratings are provisional and irrelevant, I forgot to remove them

Now, my questions :

1. Would 9.Nb5 have been crushing (as I think it would)?
2. Would 19....Nc5 have given black a huge advantage?
3. At move 23, is Bxf6 the best move? is white much better?
4. Still at move 23, ignoring Bxf6, is Ba7 Ba8 a sound plan (knowing black will try some mating attack too)? It looked really funny, but I think it is not.
5. At move 32, I played Qxd5, with only a small advantage against the proper defense (which was played), I feel like 32.Rxd5 is much much stronger. Am I right?
6. At move 34, I retreated my bishop for no reason, and should not have been able to grab the f pawn. I feel like the position went from winning (edge, but enough to enforce the win) to equal. Am I right? Am I overestimating passed pawn in such endgame? Am I missing other factors (like my passive and precarious King)?
7. During the endgame, did I mishandled Black's knight? And let him be too powerful?
8. During the endgame, could and should I have somehow activated my king more?
9. More generally, have I overestimated the bishop pair?
10. More generally, have I evaluated my King safety correctly?
11. Have I missed an attack (like 32Rxd5, or other)?