I'm still thinking about my Bg5 move. Assuming I didnt play that silly e6 move, (1) I could have taken his knight but take the risk of weakening my LCS by losing my LSB and Michael's fianchattoed Bishop gets stronger without the Knight blocking (2) I was hoping to tempt the advance of the g-pawn to weaken the pawns in front of his castled King (3) I move Bd7 but block my Queen defending my d4 central pawn or (4)as littleLizz mentioned, I simply move my Bishop back to its starting square at c8 but I'm back with an undeveloped Bishop.
Vicariously-I (2200) vs ezani (1200) Game Analysis
As a post-analysis and re-assessment of the game, I will try to play from Move 20 onwards with the Computer being Me and with Me being Michael, to see if I could really have queened one of my pawns or whether (as what happened), the Knight was able to stop it or whether the game could have been drawn. Anything except the second outcome means I didnt play accurately enough and should have taken longer time in between moves accurately calculating out possible candidate moves.

I would recommend playing h6 so your Bishop has a retreat square. Then you don't have to worry about it getting trapped. For example:
As for trading all of the pieces that is a bad idea. I went into that endgame because I knew you had no chance of winning or even drawing without any pieces to cause me trouble. If you had a lot of extra pawns then it might be different because my Knight might have trouble stopping them but that wasn't the case here.
If his Knight had taken my Bishop (assuming I didnt retreat it), my g-pawn would then eat his Knight leaving an open g-file which means I can't castle kingside and I'd have to keep my King in the center (or castle queenside?)