Is that position promising for black? Well, sure. It's also promising for white. Generally most people don't go for the exchange sac unless they've at least netted a pawn out of it first, but the key to that position is that you can't look at it as won or lost yet. You have to like the idea of playing out that endgame, or a similar, following position where black plays Bxg4 and ends up with a whole slew of pawns for a rook.

It's pretty rare that you'll see white defer castling for that long, however.

More to the point, really, though, is that black actually usually wants both rooks on the board in those endings. Simplifications help white. This is particularly true in amateur play when one less rook means one less target for a knight ending. In the position you give, simplification to a R. v. B ending looks like an easy win for white. In fact, that's his strategy.

Ultimately, I feel like the exchange sac may be slightly premature in that position. It feels a little robotic, although examining the truth of that position is beyond my ability. Nd5 isn't a threat yet (while Rxc3 is a move you should constantly be considering, I might consider Rc4 here - a different exchange sac, to net your the dark-square bishop).

The game Leygue-Rausis is given in Ward's "Winning with the Dragon 2."