This is some great analysis. I have never faired very well against the Budapest Gambit and I am definitely going to add this to my repertoire. I have actually played the Alekhine Variation before (because I've pretty much tried everything), but I never realized the attacking ideas behind it that you've explained so well. I don't face the Budapest Gambit very frequently, but I'm actually looking forward to it, now. :) BTW, at 4:33, when you have black playing 14...Qd7 because of the discovered attack on his queen, the engines (Stockfish, Leela & Komodo) all agree that black's best response is actually 14...Qd8. I imagine it is because it allows the other knight to develop to d7 instead of the much more awkward way you have demonstrated with black having to develop the knight with 15...Na6. You say, "...it is the only way to get the knight out," but that's only because of 14...Qd7. In fact, the difference in the position evaluation is rather significant. Stockfish: 14...Qd7 (+1.80) vs 14...Qd8 (+0.21), more than a pawn-and-half difference in the evaluation! Sure, white still has the advantage but +0.21 is almost negligible. Perhaps the thinking behind 14...Qd7 is that black could retreat his bishop with 15...Bf5 and, perhaps, exchange light-squared bishops which could strengthen white's attack on the king's position. However! White has a great response to that: 16.g4! If the bishop grabs the "free" pawn, the opening of the g-file will likely prove fatal.
