Agree. He also advanced the methodology of preparation for unique aspects of the game, describing a scientific, pattern recognition based approach, not simple ad hoc absorption.
Going back to Roger, I think he made the correct call on the first day Daily Double. It was a second row DD, which usually means it is going to be an easier clue. Doubling up his 10,200 would have given him a large lead heading into FJ. With such a large lead he would have not had to make a big bet on the second day Double Jeopardy Daily Double.
Not only that, but even if Roger only nailed the second DD ("Winnipeg" instead of "Manitoba") he would be on a path to winning the tournament with thousands of points to spare. All in all, he lost more than 20,000 instead of winning 20,000 on DDs - that's a turnover / flip of more than 40,000 points.
Nothing wrong with the strategy — as Ken said, somewhat wistfully, "That's what he does" — Arthur Chu has recently done the same, and both Ken and Brad were successful at employing it as well, especially when done early enough in the game, though not as methodically.
It was a game between equals. Amazing tournament!
The one downside to these tournament is the inevitable letdown of watching regular play afterwards. While Julia (the defending champ) has had a great run of 10 wins, she misses a lot of questions that I would think would be common knowledge amongst Jeopardy contestants (let alone one who has won 10 games in a row).