The only problem with the doubled pawns is that it makes it harder to deal with White's locking up the pawn structure on the dark squares.

Black's K-side pawns structure is not really "busted", Joanne. The inconveniece of doubled pawns "away from the centre" is pretty minor, and quite compensated for, in my view, by possession of the bishop pair. True, I'm not one to set a whole lot of store in having the greater ecclesiastic presence on the chessboard, but it ought not to dismissed out of hand, neither. In any case, the opening of the e-file, without the slightest effort on your part, can also be considered very convenient. Much will depend on how the rest of the game develops.  

After White's 12.Rf4 seemed a good moment to get in ...g5. In the game continuation, one feels that allowing White to exchange on e4 was more in Black's interests than taking on d2. Black's remaining bishop don't look very happy...

White's exchange on d5 looks counterintuitive, freeing up Black's bishop with no compensating advantage to White.

Possibly inaugurating a minority attack with the view of leaving Black with an isolated pawn - probably on the c-file, a weakness to aim at whilst keeping the central majority ready to advance at any time.

"Send her victorious!"

White has gone rather awry and from maybe slightly better than even 10 moves ago has rather the worse of it.

Black probably ought to prefer to hold her majority (2-1 or 1-0) on the K-side for as long as possible, and use one of the majorities to force White to give up the rook. If it comes down to a rook ending with Black's having only the b-pawn left (say), White might be able to obtain good drawing chances, unless the b-pawn has advanced a long way in the meantime.

49... Re6? leads to a theoretically drawn position; 49... Kc6 50. Rg6+ Kc5 51. Rg5+ Kc4 52. Rg4 b5 53. Kf5+ Kb3 54. Rg3+ Ka4 wins. White found the only move that holds the draw in 50. Rg1, but 51. Rb1? gave it away again. Unfortunately 51... Rc6? was the wrong continuation; only 51... b6 wins, but the continuation is too long for me to give it here.

It seems nuntar has researched this. I can't comment on 50.Rg1 as a drawing resource, but it seemed to me that the Black R ought to stay on the e-file to keep the WK out of the action. Earlier I thought that the liquidation of the K-side pawns would leave White with fine drawing chances, but thinking about the cut-off effect of the BR suggested that maybe Black's plan had its upside.