Thanks for this very informative reply. I stand guilty of sloppy research, I had simply overlooked the existence of this part 7 in CCN 94! It looks very useful, certainly it is nice to have a source that is more thorough than Palkövi and more White-friendly than Silman.
So it appears that Black can be very confident of his chances after 12.Rd2 0-0 13.Rad1, with both 13...h6 and 13...Rc8 looking promising for him? I have a lot of ideas to chew on now (12 b4!? MNb, 12.h3 Senador, 12.Nd5 Zelic, and 12.a3 Rc8 Equidistance [by transposition]), so I will
try to stay away from this thread until I have some serious analysis to contribute!
Just some quick thoughts on Senador-Nanjo (12.h3 0-0 Bg5): I like Black after your suggestion of 17...Na5!, but maybe earlier White could have tried 15.Rxd6!? Qc7 (again those double-attack ideas, with ...Nd4 threatened) 16.Rxf6 (or 16.Qd2!? Rfd8! 17.Rxf6) gxf6 17.Nd5 which at the very least looks scary for Black to face over the board.
As for 13...Ne8 the time-loss involved makes me a bit skeptical, shouldn't White be able to drum up at least enough compensation somehow? In CCN 94 the point is made (if I read it right) that by Bg5xe7 White exhanges off his good bishop for Black's bad one, but Be7 is also a typical case of "a bad bishop protecting a good pawn" (according to rebel strategic thinker Mihai Suba). When I play these positions as Black I think of Be7 as the one defensive piece that single-handedly holds the fort on d6, often allowing Black's other pieces to operate on either side of the board (maybe this is just a matter of taste though).
That's why I would love to avoid having to deal with Bg5 altogether, and I have wondered if Black can simply meet the slow 12.h3 with his own slow 12...h6, but of course White can consider both 13.Nh4 and 13.Nd5 a la Zelic in reply and... I need to analyse some more