Lukacs and Hazai's survey in Yearbook 87 covers this line in a lot of detail, so I recommend you examine this source carefully when studying the variation. 

The critical line is what was played in Ivanchuk-Carlsen, Linares 2008 (the main game of their survey): 7.Qe2 Bf5 8.Re1! (an important novelty at the time), when instead of the dubious 8...Bb4?! as in the game, Black should prefer H/L's advocated 8...Qf6!, when this unexplored position remains unclear. In my opinion White should retain a slight advantage with best play, but obviously with almost no games after 8...Qf6, this is open to debate. 

I also agree with Markovich that the 5.Re1 variation is theoretically harmless, although Black still has to be familiar with the theory as McShane-P.H. Nielsen, Hastings 2003 showed.

chk wrote on 09/28/09 at 07:20:32:


After dxe5 it's more like a Berlin ending with black still having castling rights.

Btw, Soltis in his Transpo Tricks book, gives the following line from a Vescovi game: 4.0-0 Nxe4 5.d4 a6 6.Bxc6?! dxc6 7.Qe2 Bf5 8.g4? Bg6 9.h4 "Black could have refuted that with 9...Qd7! 10.Nxe5 Qxd4, as Johannes Zukertort played way back at London 1883 (!)"

Quote:


With the e4 pawn off the board white has no kingside majority.