I may tidy this post up later today, but here's a quick recap: Gelfand got a nice advantage out of the opening (a Nimzo-Indian, Karpov Variation), but his 19.c5 was excessively academic. Rather than enjoying a clearly favorable middlegame, he hoped that in the resulting endgame with queen vs. rook, knight and pawn (plus more pawns for both sides) he'd be able to use the principle of two weaknesses to break down Anand's attempted fortress. His judgment proved mistaken, but he still had the better of play, and showed no ill effects from the disaster in game 8.
So the match is tied, game 10 is tomorrow, and - for subscribers - the materials will be available later tonight.