vonKrolock: <30.g4> A "won endgame" is in march for White, following, according to D.Bronstein, "a well prepared and calculated Opening, leading, in organic transition, to a combinatory Middle Game etc" Particularly annoying for Black is the pin in the h1-a8 diagonal...
<30...Bc8> ... when suddenly Taimanov starts a brilliant plan: he will sacrifice a Pawn to activate his living-dead Bishop and his distant majority
<32...Kf6> An important detail - restrains for the while white King's moves to 'h' passed Pawn's track
<33...a4> ! For it's practical value
<34.ba> "The immediate 34.g4 was better" (Bronstein)
A concrete line is not so necessary here as the black manouevre a3-Bf5-Bb1-Ba2-Bb1-a2 is to slow for the g5-g6-g7-g8 advance
Better for Black here would be, after 34.g4, the line 34... ab 35.ab Be6 36.g5 Ke5 37.Bd1 b5 etc - although White should win whith carefull play, Black chances of obtaining a draw - by stalemate, insufficient material or even wrong corner of the Rook Pawn (black coronation square) - were, at least, tangible
<37.Bg4> ! Pleasant
<39.g6> "!" Again quoting Bronstein, when Boles played 34.ab, he aimed at this position, but Mark the pianist had still a brilliant stroke:
<40...Bh7> !! Now both Pawns will queen, but White will not be able to exchange Queens, so the game changes to a protracted fight whith hundreds of checks for Black...
<45.Qg7> and <50.Qe6> are typical mistakes pointed out by Bronstein - first White should take the last black Pawn, and then not allow the black King to approach the 'h' Pawn
<58...b5> ! followed by b4 and, after the actual sequence, the black King will take the 'h' Pawn and reach 'c8' just in time to avoid the 'a' Pawn's coronation - ACTA EST FABULA