It has certainly been around a while.  E.g. I recall that when Spassky-Gheorghiu from Bath 1973 appeared in Gligoric's "Game of the Month," 9. Nb5 Qxd6 10. Nxd6+ Ke7 11. Nf5+ Kf8 12. b3 was supposed to be a strong novelty.

Reverse wrote on 11/17/16 at 14:57:23:

I dare say that if that were the case, it would be one of Black's main defenses ... 

edit:  I see that Lakdawala in fact has a whole chapter on 6...h6, and seems to portray it as enabling Black to equalize.  One line which attracted my interest is 7.Nd6+ Bxd6 8.Qxd6 Qe7 9.Qxe7+ Kxe7 10. Be3 d6 11. f3 Be6 12. 0-0-0.  As Lakdawala notes (and as I recall), once upon a time that was given as leading to ±.  But is += (Nunn) really not accurate?  ...

Cyrus Lakdwala mentions a Ulf Andersson Sicilian in his upcoming book on the Sveshnikov Sicilian: 1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 e5 6.Ndb5 h6 7.Nd6+ Bxd6 8.Qxd6 Qe7 9.Qxe7 Kxe7.

Is this a serious line? Does black equalize? What is the appeal of this line. I'm just curious...looking for an offbeat sicilian to add to my repertoire to play against lower-rated players.