10. Qe2 is "Sturm-Schmid" variation according to Harding

Before reading Watson's book, I wonder why 11.Qh5+ is a good move or not, after Keres 10...f6

Qh5 is a dubious move:  11...Kd8!  (11...Ke7 12.Be3! Qxd3 13.Bc5+ Kd8 14.Bxf8) 12.Qf7 Ne7 13.exf6 Qxf6 14.Qh5 g6 15.Qe2 Nf5 16.Bf4 Bg7 17.Nc3 g5 18.Bd2 with Watson assessment: unsafe position of Black king makes the balance with the deficit of two pawns

10. Nc3 develops quickly a piece (one of the main aims of this gambit) and also aims at:
a) Nb5 (attacking the Queen)-Nc7+ or similar, or
b) Nxd5 if the e6-pawn is pinned (i.e. after 10. ... Qxe5 11. Re1).
The Knight clearly belongs to c3, while the Queen may be better at f3 (supporting Bf4), or g4 (attacking g7), etc. in some variations. Moreover, White actually prefers (at least I!) to lure the Black Queen to take the 2nd pawn as he gains good attacking possibilities despite the pawn deficit.

Now, 10. Nc3 a6 11. Qe2 has similarities to 10. Qe2 in that it may result in a more positional game. White will play positionally (despite being a pawn down!), his pros being:
a) his greater space,
b) slightly better development (the Black Q will be harassed again), and
c) the use of d4 as a useful pivot point for his minor pieces.
The e5-pawn needs to be carefully defended (e.g. f4 may be needed) and gradually there could be prospects for a K-side attack..

Now, in case you are familiar with the strategic points I've made above and you only need some concrete lines, the most I could find is:

10. ... f6!? (suggested by Keres) 11. exf6 Nxf6 12. Rd1 Qh4; or 12. Nc3 Bd6!? 13. Nb5 Qe5 (Watson)

This looks quite different than the usual 10. Qe2 a6 11. Nc3 or 10. Qe2 Ne7 continuations, but I have not seriously analysed it to see if it works out well (and if it is the refutation?!?!?! )

Hope this helps..

White can gambit only one pawn with: 7...cxd4 8.Qe2 (8.Re1; 8.a3) and after 8..Nge7 9.cxd4 Nxd4 10.Nxd4 Qxd4 11.Nd2 (with the idea Nf3-Harding)

Just consider two safe winning lines for black. Both involve taking one pawn and returning it and liquidating into advantageous ending. 

1.e4 e6 2.d4 d5 3.e5 c5 4.Nf3 Nc6 5.c3 Qb6 6.Bd3 cxd4 7.cxd4 Bd7 8.0-0 Nxd4 9.Nxd4 Qxd4 10.Nc3 a6 11.Qe2 Ne7 12.Kh1 Nc6 13.f4 Nb4 14.Rd1 Bc5 15.Bxa6 Qf2 16.Qxf2 Bxf2 17.Be2 

1.e4 e6 2.d4 d5 3.e5 c5 4.Nf3 Nc6 5.c3 Qb6 6.Bd3 cxd4 7.cxd4 Bd7 8.0-0 Nxd4 9.Nxd4 Qxd4 10.Nc3 a6 11.Qe2 Ne7 12.Kh1 Nc6 13.f4 Nb4 14.Rd1 Bc5 15.Bxh7 Qf2 16.Qxf2 Bxf2 17.Bb1 0-0-0 18.a3 Rxh2+ 19.Kxh2 Rh8+ 20.Bh7 Rxh7# 

Line stops with move 17. Added the balance moves for the fun of it

Quote:


I thought that’s the safest, no risk way to handle Milner – Barry. Once you introduce …Nb5 before …Bc5, I think black reaches the winning ending. 

1)      For starters, black has a connected passed d pawn.
2)      Taking back either h pawn or a pawn have their down sides for white. For example, taking the a pawn often results in surrendering the Bishop for 2 pawns and taking the h pawn, (objectively better for white) results in a difficult ending for white.

My 2 paise

I think I gave variation I prefer...
Yes, if White tries Rd1 at some stage, Black has the idea of Qf2 trading queens...and it's good for Black. Playing an endgame without queens and a pawn less is no fun for White.
One typical variation is given in NJ Chess forum byPete Tamburro, with the comment: the old Rd1 was out of fashion after the famous "Potter memorial" tournament.
10.Nc3 a6 11.Qe2 Ne7 12.Rd1 Nc6! and so on (I will continue only in one specific thread) or 12.Rh1 Nc6 13.f4 (13.Be3!? Bohak attack) Bc5 (13...Nb4 14.Rd1 Bc5!? Piskov) 14.Rd1 Qf2 and Black is clearly better.

Dom thanks for the reference to the French website!