[Bongle]

I agree with all your points, but I think it misses one point. I'll try to write it in your style:

The game should have fairly consistent movement
Games where the robots tend to bunch up tightly and stay there for extended periods are boring. Also, they tend to result in broken robots. Games with more open floors or fewer chokepoints tend to have a greater amount of robot movement, which makes for more exciting games and a feeling of flow.

Good Games from this standpoint:
2006: The period changes always precipitated a big race from one end to the other as teams rushed to get into scoring position. Once there, there would be pushing matches, but there was usually enough freedom of movement that the defender and defendee would move around.
2008 (kinda): The very nature of the game required fast movement. However, since some robots had to stop in order to hurdle and drivers couldn't see around the overpass, traffic jams were common in the home zones.

Bad Games from this standpoint:
2007: The rack was a difficult thing to score on without being defended. Once a robot was in front of you trying to prevent you from scoring, long periods of dead time would occur as you fought for position. This also resulted in lots of broken robots, which would then become defender robots.
2009: The lack of anti-pinning rules, though good in some respects, allowed whole-match pinning as a viable tactic (which we successfully used in one match). Perhaps at IRI they should widen the carpeted portion of the field to allow pinned robots a better chance of escaping. Another problem: the addition of 6 robot-sized trailers created a situation where there simply wasn't much space.