just had a quick glance why not rook takes pawn?
Max Lange: 9...g6

Hey D.
11...Rxf7 doesn't work as 12. Qxf7 then Qh7
13...Rxf7 14. Qxf7 and that looks winning for White
14...Rxf7 (I'm assuming it's this one you're referring to) 15. Bxe7 Rxf3 16. Bxc5 looks similar to the other line, but White is now hiting at the d4 pawn.\

I'll have to look at those lines a little later, though I forgot to mention, the article is http://www.chesscafe.com/text/abby06.pdf lotsa good stuff in there!

you know after the line Bxc5 you gave with d4 dropping its still quite tricky.. for example the rook can retreat from f3 to f7 or even f5...and whites weak back rank forces Na3 which may not be clear to your opponent..as a result i think its not even a full pawn ..i checked it out w a pc after i though t of this. rd8 and c5 is on the table after exchanges on d4

Interesting, I was also unaware of this line... and I used to play this before, though via a different move order.. (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.0-0 Nf6 5.d4!?)
anyway, I don't think the transition to the endgame is forced, though white is probably no more than equal here. If you don't want an endgame, here is an idea I've had - don't play Qf3! (even though computer says it's better :) ) I prefer moving the knight, as it seems more tricky and here are the lines that I've checked with my engine:

8.fxg7 is stronger, although traditionally less commonly played, and white gets a slight edge after a load of complicated theory in my study and Black's position is quite unpleasant to play if White is well prepared. If I could always get the Max Lange attack, I would play for these positions more often (instead of playing 3 Bb5)! For instance if Black plays 3...Bc5 followed by ...d6, I do not recall the line but I don't think White much to play for objectively. Marin's book, Beating the Open Games, goes into these lines in detail, a lot of the theory in these lines are not widely known. I just bought the book last week so I haven't read it, but it looked excellent for a player looking to counter 1 e4 with 1...e5. There's a lot of theory, and it's very easy for White to play given that he knows the theory. It's a particularly useful weapon to play and beat masters. The main drawback is that analyzing playing these positions does not increase understanding of chess as much as a Ruy Lopez, and White has sacrificed a pawn, although as far as pawn sacrifices go, this is one of the most critical ones.
here's a win for me, at one point conveniently Black took on g7, making it much easier, instead of playing ...f5, maybe best, or ...d3 although at the time I was well prepared, it would have certainly been a much more difficult game

Thanks for the game, finally getting around to this and the line you suggested, I tried it in a quick game a few months ago against a local master, I remembered the 8. fxg7 idea, but I played something a little different after and was surprised at the line that occurred, we both got mixed up with the traditional main line and Black was quickly busted, if I recall, it went:

Some like to relive 'the good ol' days' of chess is my guess :)
Actually, against Nxe4 I've been following some recommendations by Stefan Brucker in an article on Chess Cafe 'A rook with a view', the idea is instead of Bg5 to play Ng5 and play against a backward e pawn for a bit.

It seems like analysis really likes the max Lange attack for white. It looks really hard to deal with..maybe taking on d4 with the pawn just isn't best??
I was going through a ChessCafe article by Abby Marshall, in one line, she gives the following variation:
Am I missing an important variation after 9...g6 that White can play to avoid the resulting endgame?