Pawn down queenless middlegame, with adequate compensation (bishop pair, slightly better development). Analysed a lot, without any concrete plan for white. At least mr. Radjabov has played many times this opening against top-notch competition, and has never lost a single game!

You don't have to understand 12...Ke6. This is run & gun chess, only engines at deep ply can manage that adequately. But as I said, the whole 5...d5 variation is under a heavy cloud because of 6.Nxe5 de4 7.Nxc6 Qg5 (there is also 7...Qd5, although white can get an advantage here as well). 8.Qe2 Nf6 9.Nxa7+!

Isn't that adequate? How come?

In Chesspub a few opening experts tried to work out an advantage for white in the tabiya after 6.Nxf6+ Qxf6 7.Qe2 Be7 8.Bxc6 bc6 9.Qxe5 d6 10.Qxf6 Bxf6, without success. Mickey Adams tried this OTB a few months ago against Tomi Nyback, without success. And a 2660 rated player lost as white against Azarov.

Suffice to say that white's most meaningful attempts for an advantage are certainly not "obvious" looking moves. But apparently, you know better, even if you have not given the position even a casual look...

If you play this position against someone "your rating" then it's absolutely certain that the one who blunders last, will lose.

But this is totally irrelevant to the opening.