The reason Watson stopped recommending this line is because of the sharp variation
1. e4 e6 2. d4 d5 3. Nc3 Bb4 4. e5 Qd7 5. a3 Bxc3+ 6. bxc3 b6 7. Qg4 f5 8. Qg3 Ba6 9. Bxa6 Nxa6 10. Ne2 Nb8 11. Nf4 Nc6 12. Nxe6 Qxe6 13. Qxg7 Qg6 14. Qxh8 O-O-O e.g. in Albano - Horwitz, corr USA 1987
https://chesstempo.com/gamedb/game/2330594 .
The moves
15.O-O,
15.Be3, and
15.Kf1 were extensively analyzed in
Inside Chess, by Mike Valvo (1988.10.09), by Ian Rogers (1989.04.03), and by John Watson (1989.05.29). Since that time
15.Bd2 has also been played.
Quote:What a mess! Offhand, I don't see what's wrong with using black's ...h6 theme earlier, i.e., 17... h6. ... I no longer trust any of these positions for either side!
--Watson
Watson's ...h6 is referring to
15.Kf1!? Rd7 16.h4 Rg7 17.g4 (Callaghan - Blechar, corr CCLA 1985) and now
17...h6.
I notice that Seirawan was still willing to go into this as black, e.g. in Kamsky - Seirawan, Tilburg 1990
http://www.chessgames.com/perl/chessgame?gid=1066577 , and Fedorov - Seirawan, Istanbul Ol 1990
http://www.chessgames.com/perl/chessgame?gid=1402465 . Also Hiarcs Master 3.0 - Seirawan, AEGON 1995.
Probably Lakdawala is going to recommend
5...Bf8. White I think can follow Kasparov - Ivanchuk, Horgen 1995
http://www.chessgames.com/perl/chessgame?gid=1060586 . Keene's annotations at chessgames are a little misleading. According to Hodgson in
CHESS Monthly (1996.January), Kasparov was doing well in the opening, but was "not on the ball" due to illness.