"I don't think people realize just how strong players of his caliber are. I'm sure that against a 1700'ish player he could clog up the board, win a couple of pawns or the exchange while some minor material is being traded, and then apply enough pressure to at least get a draw 50% of the time, and probably a win 30%+ of the time."
I don't think that's true at all, it might not be that hard to put pressure on a weaker player when you are starting on equal footing, but when you are a rook in arrears to begin with, it's not that simple. Applying pressure to win a couple of pawns is certainly not going to overcome the rook you are behind even if it were. Granted I'm not a 1700 player anymore, but I'm quite sure when I was, I would have beaten curtains at his present strength 95%+ of the time with rook odds.
"Use of that extra rook often doesn't come into play until the end of the middle game, after some files have been cleared/pieces have been traded."
That's completely wrong, and I frankly can't comprehend how you could even begin to think that's true even if you're a fair beginner at chess. The concepts of opening, middlegame, and endgame can hardly even be used in such a game, rook odds skew the game far too much to properly define what those really mean in an odds game.
Ok but adanthar is using his yahoo rating which is completely meaningless. This is why there are so many people who think they are better than they aer, because they believe these random Internet chess sites to have somewhat accurate ratings. You can easily be 1700 and actually play at about a 1000-1100 level.
Honestly if you've never played a real chess tournament, its 95+% likely that you are gigantic underdog against me if I give you rook odds.