If you push your pawn too fast up to b4 it can turn out to be a weakness and after a hypothetical a3 by White you may have to play bxa3 (bad).
As Bronstein said, the knight wants to go to d5 anyways, instead of blindly pushing queen-side pawns it makes more sense to prepare the d5 push with the maneuvers he mentioned.
Those are my 2 cents anyways, I hope an experienced player can clear things up for the community.
Thanks, Che. That does make sense, and clears up the basic confusion. And yes, I do want more input, from players of all levels of experience - I put this here for a forum discussion about the whole opening strategy in this position, to see what all others notice in this position - I would have seen nothing, without Bronstein's explanation (and yours). So thanks for the input 
You are welcome, to keep the chess discussion going, I learnt the early b4 answered by a3 "motif" by studying the Closed Sicilian. When someone says "Closed Sicilian" the first thing that comes to your mind is White sacrificing all his pieces on the king-side however it is not like that, White also pays a lot of attention to the queen-side.
In this position, after the moves given, Bronstein, in his famous book of the Zurich Candidates Tournament, writes: "With the white king's bishop fianchettoed and the black e-pawn advanced to e5 in this variation, the moves 5...a6 and 8...b5 are not intended as part of a queenside attack (since there's nothing there to attack), but rather to gain space for the black pieces: ...Bb7 and ...Nb8-d7-b6. From this point of view, the bishop's deployment at e6 makes little sense." (game 66, p104, Gligoric - Kotov). What does he mean by this? Why is there nothing to attack? He has, after all, the queenside pawns, so getting eg. the knight out of the way with ...b4 surely opens that up. Bronstein does some further explaining, that the knight will only be driven to d5, but surely that gets it out of the way for the sake of an attack on c2? What am I missing?