Ok, so another interesting tidbit is the comparison between calculations on Earth in January with 1) a starting ice cap on Antarctica, and 2) without a starting ice cap. When the model starts with ice, things look reasonable enough, but when it doesn't... the south pole gets hotter than the outback in the summer. This seems like a spurious result and, even if it isn't, does go to show the massive (50 degrees C!) dependence on the starting conditions.

With Ice (Left), Without Ice (Right)
earth_jan_ice.png earth_jan_noIce.png