5. d5 does NOT win a piece, 5. a6 counters the pin. The 'pressure' on the knight you discussed is pretty phony, as the simple 5. Bd7 relieves the pressure. It's still the best variation for white though, maintaining development and tempo, without creating any weaknesses, blah blah blah...

Castling here also seems good. Personally I would have left the bishop on b5 until Black played a6 but I don't see anything terrible or bad about 4. Bxc6+

I probably would leave the Bishop too now, I chose this game because of how long ago it was. Mainly it was to see if I could see mistakes and ideas that I missed. It was an educational experience. I've found that two years ago I liked to break up the tension all the time, but now, tension is a good thing (unless I'm on the receiving end of it ;-P).

There's nothing wrong with this variation either, all the book moves at this stage are sound, it's just a matter of picking the one that takes to your fancy most...

Of course... Be3 is better here, now you'll have to move the knight again, or play b3, to reactivate the bishop. It takes less time to play Be3 first, then play Nbd2 if you wish.

Just take a bit of time to think and you won't miss golden opportunities like that man... just think 'i've stopped the checkmating threat, so now the opponents e pawn is threatened.' Easy picking...

true, true... Bd6 should be forced for black here, but he decides to toss away the e pawn. Silly...

You could have made life easier with Nxe5 on 11. Wins the pawn, which has already used one get out of jail free card. Black would then try and defend c6 with Bd7 and you can continue your knights powerhouse plan with 12. Nc4 Bb6 13. Bg5, which will really start to screw in the screws in blacks coffin.

Your c3 move has helped set this all up. At that point, your were actually thinking, which is handy, but your not spotting the opponents mistakes, possibly because your concentrating on your own gmeplay not the opponents. You should take equal consideration of them both...

e5! Is much simplier and wins a piece.

Good point! It seems I must have gotten so wrapped up in my plan that I forgot to look for blunders like that! And apparently in my analysis too! I'll have to fix that.

18. Nxg4 is followed by 19. fxg4, giving the bishop the backup it needs.

Again, I'm glad you've caught my logic flaws. My whole thought process was revolving around hxg4. Maybe when I was playing the game I saw this... I'll have to fix my annotation again.  

seems a complete waste; in fact it drops a pawn after 19. Ne5. i don't think there's anything effective for black here. As well as being a pawn down, black has a damaged pawn structure on the c file, and white has a passed pawn and plenty of options, like Ne5, and the ability to easily get the rooks involved.

Amazing! This move doesn't solve the problem (which didn't exist) it creates it! Now black can play Nxg4 or Nxe4 (Nxe4 is better; removes the passed pawn) if he plays Rxd1 first.

Oh no! My analysis seems to be all based around this ghost! And mistakenly so! Ok, I'm going to fix all of this.

naturally, black misses the chance...

Not good enough for black if he wants to get anything out of this, it's time to drag the king into the ring with Kf7...

Surprisingly it doesn't cost black a pawn straight away, if 25. Rxb6 is played Nd7! Wins the rook for the knight (defending the bishop with 26. Re6 fails to Kf7.) Certainly though white will be able to win it soon, so definitely time to resign for black...