Aronian - Karjakin: Rules are for Fools

True masters of their craft regularly ignore rules and recommendations that act as clutches for the rest of us. Consider the following position from a blitz game between Levon Aronian and Sergey Karjakin at the 2015 Zurich Chess Challenge after 19. ... Rb8

White (Aronian) to move after Karyakin's 19. ... Rb8

Aronian played 19. Kf2 and eventually won the game after both players put their kingside pawns in motion.

I'm wondering though... in the diagram above... Why didn't Aronian just take the bishop on b7? 

It's a well-known rule of thumb that in endgames with pawns on both wings, the bishop is superior to the knight. In the diagram above, there may even be the chance to lock Black's queenside pawns in place on light-colored squares, which would make White's bishop even stronger.  

Obviously, Aronian is aware of all this, so he must have had his reasons to keep his knight (and Karyakin's bishop) on the board.

Would this not be a textbook example of a position where the Bishop is stronger than the Knight?

Was he afraid that the afraid that the advantage wouldn't be enough to win the position?
Did he see "ghosts" or overlooked something? After all, this was a blitz game. Or is the evaluation of the diagram above different at a super-GM level compared to mere amateur play?

Aronian won the game regardless. However, Karyakin "helped" by pushing for a win himself. It seems to me that if Black chooses to play for a draw, the task is much easier in the first diagram than in the second.

The complete game can be found here:
Aronian - Karyakin, Zurich Chess Challenge Blitz, 2015

I really don't know. Which is why I was fascinated by this example and chose it for this blog.

Oh well... these guys are 2700 GMs for a reason, and I am not...

Bishop and Pawn vs. Bishop

This is a position from a practice game against one of the people I'm giving chess lessons to. My opponent resigned a little prematurely. The remaining moves would have been quite instructive: