Yes, the Chess Informant GM panel just awarded the Best Game Prize for Chess Informant 97 to this game, so they agree with you. In his notes to the game, Kramnik writes 26 xf7+!! so he clearly agrees with you about the difficulty of spotting this move. Personally I thought Kramnik's win over Aronian was much more impressive and that Bruzon played this game rather poorly.

Position after 26 xf7+!!:

click for larger view

Kramnik is using an idea found in an opening trap in the Modern Defense.

The d6- and e6-squares (Q6 and K6 in Descriptive Notation) are the two squares that the White knights fantasize about occupying. Now consider the e6-square in particular. In the starting position it is <GUARDED> times, by the Black f7- and d7- pawns and by the Black c8-bishop. But Black moves his d7-pawn away from <GUARDING> the e6-square on move 1 (1 ... d7-d5) and then <BLOCKED> his c8-bishop with his b8-knight on move 5 (5 ... b8-d7). So after move 5 Black has only -one- defender of the e6-square, the Black f7-pawn.

So with the sequence 26 a2xf7+!! g8xf7 27 d2-a2+ f7-f8, Kramnik <REMOVES THE GUARD> of the critical e6-square by first <CAPTURING> the Black f7-pawn which defends it and then <DRIVING OFF> the Black f7-king which replaces it. The result of this brilliant tactical sequence is that the e6-squared has been cleared for occupation by the White d4-knight with 28 d4-e6+.

Position after 26 a2xf7+!! g8xf7 27 d2-a2+ f7-f8 28 d4-e6+:

click for larger view

With 15...e5, Black tries to reduce the dynamic potential of White's pawn phalanx by creating a strong point in the centre - a concept that has existed since the time of Stenitz. With his following moves, he over-protects his strongpoint - a concept that has existed since the time of Nimzowitch. Should White play dxe5, Black intends to recapture with the knight and establish a blockade. This concept has existed since the 19th century.

One thing it does is leave Black cramped under the strong White center, which may have led to 15...e5 as an attempt to gain some space. Certainly, as <newzild> points out, there are some good reasons behind the move. Bruzon wouldn't have tried it otherwise.

But there are problems as well. It does make the final attack along the a2-g8 diagonal possible, but more immediate is the obligation it imposes on Black's pieces.

Of course Kramnik doesn't play dxe5, but maintains his center. Now the Black d7 is kind of stuck. If it ever moves, White plays Nxe5, winning the pawn while retaining his center. But if the knight is stuck on d7, than the c8 can't play. Black's play may seem aimless and time wasting at that point, but I don't see a lot of good alternative plans in the kibitzing.

This game reminds me a lot of some of Anderssen's play in the Evans Gambit. No hurry; just maintain a strong center and build up the attack behind it. One example is Anderssen vs Zukertort, 1869. Some of Anderssen's Evans Gambit games look like he's play the Ruy Lopez a pawn down.

click for larger view

and White still has work to do. The game line 30 Re3 is now met by 30...Nde5! since 31 Rd8+ is no longer checkmate.

Enjoy finding the win! Crafty EGT link to the position at move 29: http://www.chessvideos.tv/endgame-t... You are White: drag and drop the move you want to make starting 29...Qxe6.