Kasparov lost game two the other day, but today he apparently gave X3D Fritz quite a beating in a closed position, which is where positional play tends to be more important than tactical computation. This doesn't mean that the computer is broken, or even that Kasparov will win, but it doesn't seem to be the death knell for competitive chess.
I find this whole idea that chess will be over once computers beat the best humans regularly a little confusing. It's not about finding a solution to chess, because there's no solution in sight even with computers, and in any case they provide an excellent vehicle for continuing the search. It doesn't even seem like an undefeated computer would put chess in the realm of the purely computational, since humans with over the board experience are programming the machines with axiomatic positional expertise. Is it just a matter of pride?