Also reminds me of Richard Lindzen's Iris hypothesis, which suggest that temperature increases in the tropics will reduce cirrus clouds and thus let more heat escape from the atmosphere.
Lindzen and his theories are controversial, but I do think there may not have been enough focus on possible negative feedback mechanisms like this.