"C.O.O.L. Chess" by scottish GM Paul Motwani ...

9.b3 oo 10.Bb2 b5!? is Bologan's suggestion

Don't be afraid by one variation, chess is made of improvments and work/study is allways useful..

Yes, the plan of Ceg5 and an attack is one plan for White, but if you are well prepared, you can play the line and improve next moves or choose another line.

First, for the sacrifice: 6...Nd7 7.Nxf7 Kxf7 8.Ng5+ Ke8 9.Bc4 (9.Nxe6 Qe7 10.Qe2 Kf7 11.Kg5+ (11.Nxc7 Qxe2+ 12.Be2 Rc8 unclear) Ke8 12.Ne6 draw) Bxg2 10.Rg1 Bd5 11.Nxe6 and now instead of 11...Bb4+ Kotronias-Franchini,Reggio Emilia 2005, 11...Bxe6 12.Bxe6 Qh4 Sorensen-Hillarp Persson, 2000 (line given in the forum threads)

Second, if you fear the sacrifice then 6..Bd6 or 6...Be7 are your moves.
The idea behind Bd6 is to control e5 always useful here because Black must forbid Ne5 at some stage, and for 6...Be7 a quieter move, Black controls the g5 square.

6...Bd6 7.Bd3 h6 (the logical move, Black can't play h6 one move before: 6...h6? 7.Nxf7 Kxf7 8.Ne5+ +-) 8.Ne4 Bxe4 9.Bxe4 c6 Anand-Karpov, Linares 1993
or 7.Bc4 h6 8.Nxe6 fxe6 9.Bxe6 Qf6 the right square for the queen Aronian- Turner Hastings 2000
or 7.Nxf7 Kxf7 8.Ng5+ (8.Ne5 Bxe5 9.Qh5+ Kf8 10.Qxe5 Qd6 and advantage to Black) Ke8 9.Nxe6 Qf6 (again, the good square) and Black has good game, you can check variations with Fritz. 

All the ideas exist in other variations, thus it's useful to continue the study of the lines. Example: 6.Bd3 Nd7 7.oo Ngf6 8.Neg5!? Bd6! (8..h6? 9.Nxe6! a sharp sacrifice but not so easy to find good moves for White 9..fxe6 10.Bd3+ Ke7 11.c4! (11.Re1 Nb6 (11...Bxf3 12.Qxf3 +- ; 11..Bd5 12.b3! ) 12.Ne5 Be8 unclear) and White has good pressure on Black position) 9.Re1 (9.Qe2  Large-Hodgson,Londres op 1989 (Psakhis)) h6 Harley-McDonald,Cheltenham 1994 

I took another look and I think that some of those Fort Knox variations may be pretty dangerous for black.  On the Chesspublishing site, Neil McDonald gives the following game:

http://www.chessgames.com/perl/chessgame?gid=1384388

The game follows the same move order as line A3, except that black plays 7. ...Nd7 instead of 7. ...Bd6

I'm not sure, but it looks like white can still generate a very dangerous attack with 8. Nxf7 Kxf7 9. Ng5
against 7. ..Nd7.

It looks like black can probably remaind ahead with perfect defense, but it still looks dangerous.  Is this a critical line?  Are there any improvements for black?

My suggestions for less theorical work: Fort Knox variation  and Katalymov's variation (= neo-Rubinstein).

I think Neil McDonald will vote for the Fort Knox, if he is aked for a good (and solid) variation with good reputation. I recall he earned some of his norms with it.

Here are some lines but you can collect many ideas searching in the updates and the forum.

Fort Knox start with: 1.e4 e6 2.d4 d5 3.Nc3 (Nd2) dxe4 4.Nxe4 Bd7

A) 5.Nf3 Bc6 (5..Be7 is a good alternative) 
A1) 6.Bd3 Be7!? (6...Nd7 is considered as the main move, because control of e5 is often mandatory but here the d4 pawn is attacked) 
A11) 7.oo Nf6 8.Ng3 (if 8.Qe2 Black will exchange some pieces on e4 square and will have a solid middlegame) Qd6 (or 8..Nbd7)
A12) 7.Neg5 Nd7! 8.Qe2 h6 9.Nge4 Ngf6 10.Nxf6 Bxf6 11.c3 oo 12.Bf4 Re8  Westerinen-Noegueiras,Yurmala 1978 (Psakhis)
A13) 7.Ng3 Nf6 8.oo see A11
A14) 7.Qe2 Nf6 8.oo see A11
A15) 7.c3 Nf6 8.Qe2 Nxe4 9.Bxe4 Bxe4 10.Qxe4 Qd5 11.Qxd5 exd5 = Pavasovic-Morozevich,Plovdiv 2003
A16) 7.Bf4 Nf6
A2) 6.Qd3!? (Smyslov) Nd7 7.Be2 Be7 8.c4!? Bxe4 9.Qxe4 c6! 10.oo Ngf6 11.Qc2 oo 12.Bd2 Re8 = (Psakhis)
A3) 6.Neg5!? Bd6 (6..Be7 7.Bd3 Nd7 transposes to A1) 7.Bd3 h6 8.Ne4 Bxe4 9.Bxe4 c6 Anand-Karpov, Linares 1993
A4) 6.Ned2 (Universal system, with the idea Nc4 and control the e5 square) Nf6 7.Bd3 Nbd7 8.Nc4 (8.oo Nd5! ; 8.c3 Bd6 ; 8.c4 Bxf3 9.Nxf3 (9.Qxf3 c5 10.Qxb7 cxd4 11.oo Nc5 12.Qf3 Nfd7=) Be7 transposes to A1) Kobalia-Shaposhnikov,Baklan 2005 (Chesspusblishing september 2005)

B) 5.Bg5 Be7 6.Bxe7 Qxe7 7.Bd3 Bc6 8.Nf3 Nf6 9.Nxf6 Qxf6 Shabalov-Anastasian,Minsk 1990 (Psakhis)

C) 5.c4!? Bc6 6.Nc3 Nf6 7.Nf3 (7.Be3?! Boudre-Apicella,Val Maubue 1989 (Psakhis)) Bxf3 8.Qxf3 c6 Karpatchev-Foisor Ovidiu, Bischwiller 1999

D) 5.Bd3 Bc6 tranposes to A) lines

**************

The second suggestion is neo-Rubinstein stating with 4...Qd5!?

A) 5.Nc3 Qa5 6.Nf3 c6 7.Bc4 Nf6 8.oo Bd6 or 8.Bd2 Qc7 9.Qe2 Nbd7

B) 5.Bd3 Nc6 6.Nf3 Nf6 or 6.c3 e5 7.Qe2 

C) 5.Qd3 Nc6 6.Nf3 Nf6 7.Nxf6 gxf6 8.c3 Bd7  Solleveld,M-Ernst,S/Leiden 1997 or 8.c4 Qh5 9.a3 Bd7 (Dom, the idea is to delay the Rg8 move)
or 9...Rg8 10.Be3 Bd7 Horvath,A-Hoang Thanh Trang/Hungary 1998

You should ask yourself what kind of position you want to play. If you don't mind to give up the centre and have somewhat passive play, looking out for a counterstrike, then the French Rubinstein is a good choice.
There is another way to cut down theory as Black: use transpositions. After 1.e4 e6 2.d4 d5
A1)3.Nc3 Nf6 4.Bg5 Be7 5.e5 Nfd7 6.Bxe7 (6.h4) Qxe7 7.f4 a6 8.Nf3 c5 9.dxc5 Nc6 10.Qd2 Nxc5 and 11...b5.
A2)3.Nc3 Nf6 4.e5 Nfd7 5.f4 c5 6.Nf3 Nc6 7.Be3 a6 8.Qd2 b5 9.dxc5 Bxc5 10.Bxc5 Nxc5 and 11...Qe7.
B1)3.Nc3 Nf6 4.e5 Nfd7 5.Nce2 c5 6.c3 Nc6 7.f4 cxd4 8.cxd4 Qb6 9.Nf3.
B2)3.Nd2 Nf6 4.e5 Nfd7 5.c3 c5 6.f4 (6.Bd3 Nc6 7.Ne2 or 7.Ngf3) Nc6 7.Ndf3 cxd4 (Qb6 and a little later cxd4) 8.cxd4 Qb6 9.Ne2 (9.g3).
Still it's more theory than the Rubinstein of course - and a completely different type of play.