According to Stockfish6 at depth of 30 10.Bf4 is busted by 10...h5.  Sorry for reviving an old thread but I did start it in the first place, plus I want somewhere to keep my notes for an upcoming thematic Winawer tournament on chess.com that includes an IM (http://www.chess.com/tournament/french-winawer6)

10.Bf4 h5
(10...h5 Stockfish 6 64 -0.98 (depth 31)  11.Qh3 a6 12.Nd6 Bc7 13.Nxc8 Qxc8 14.Be2 Nf6 15.exf6 Bxf4 16.fxg7+ Kxg7 17.O-O e5 18.Qxc8 Rhxc8 19.g3 Bh6 20.Rad1 f6 21.Rfe1 b6 22.Nh4 Ne7 23.Bxh5 Rxc2 24.f4 Rh8 25.Bg4 Rc3 26.Rxd4)

(10...h5 -1.08 (depth 31)  11.Qh4 Qxh4 12.Nxh4 Nge7 13.Be2 a6 14.Nd6 f6 15.exf6 gxf6 16.Bg3 e5 17.f3 Kg7 18.Bd3 Bc7 19.Nxc8 Nxc8 20.Nf5+ Kf7 21.O-O Nd6 22.Nxd6+ Bxd6 23.Rfb1 h4 24.Be1 Ne7 25.b5 a5 26.a4)

(10...h5 -1.26 (depth 30)  11.Qg3 h4 12.Qh3 Nf6 13.Nxh4 Kg8 14.exf6 Qxf6 15.Qg3 d3 16.O-O-O a6 17.Bg5 Qa1+ 18.Kd2 Qb2 19.Bxd3 axb5 20.Rb1 Qe5 21.Bxb5 f6 22.Bxc6 Qd4+ 23.Kc1 fxg5 24.Nf3 Qxf2 25.Qxf2 Bxf2 26.Bb5)

(10...h5 -1.27 (depth 30)  11.Qg5 Qxg5 12.Bxg5 f6 13.exf6 gxf6 14.Bd2 Nh6 15.Bxh6+ Rxh6 16.Nd6 e5 17.Nh4 Ne7 18.Nxc8 Rxc8 19.O-O-O f5 20.Bb5 a6 21.Ba4 e4 22.Kb1 f4 23.Rhf1 Kg7 24.Bd7 Rd8 25.Bh3 Rc6 26.Nf5+)

You're right!

In fact ICCF has quite a good gameviewer, so the easiest thing is to give the links. The two games are available at

http://www.iccf-webchess.com/MakeAMove.aspx?id=435090

and 

http://www.iccf-webchess.com/MakeAMove.aspx?id=434549

There are two correspondence games from the latest ICCF games archive (server month 201309, released today, I think) in just this line (to 14 ... Bxd8):

Lil'-Schott, Chess Planet-DESC ICCF corr 2013 (1/2-1/2, 44),
Enin-Schott, Bestlogic-DESC ICCF corr 2013 (1/2-1/2, 43).

ICCF requires signing up to access their archive, so I think I need to respect that and not upload the games;apologies. But I've found it well worth signing up, and recommend it.

In the major threads I've found on the Armenian variation over the past several years, I've not seen 7...Kf8 mentioned as an alternative to 7...Ne7
in the line of the Armenian that goes 1. e4 e6  2. d4 d5  3. Nc3 Bb4  4. e5 c5  5. a3 Ba5  6. b4 cxd4  7. Qg4.  However I have found a game where
Lputian employed 7...Kf8 (against Firman in 2005) but I think after 8.Nb5 Bb6 9.Nf3 (9. Nxd4? Qc7; 9. Nd6?! Nc6, ...a6 and ...Bc7) Nc6 10.Bb2 his 10...f6 might have been too reckless as his opponent missed 16. 0-0-0 after the sequence 11.Nbxd4 Nxe5 12.Nxe5 fxe5 13.Nxe6+ Bxe6 14.Qxe6 Qf6 15.Qxf6+ Nxf6 instead settling for 16. c4 and the game was eventually drawn.  But instead can't Black get the queens off more safely after 10...Nge7 11.Nbxd4 h5 12.Qf4 Ng6 13.Nxc6 Nxf4 14.Nxd8 Bxd8 and isn't this safer than the game Black gets starting with 7...Ne7