Slight quibble - the global temperature anomaly maps would be better if you gave us the temperature scale, and on my combination of wet string and MS IE6 they came out on top of the caption. Even better than the color scale (for those of a numerical bent) would be a summary of (say) max, median, min and quartile delta-T for each scenario - if you've got that handy that is.

My non-meteorologically trained gut feel for this sort of thing is that hurricane intensity should be a function of (lateral? vertical) spatial temperature gradient in the hurricane belt rather than absolute temperature. Are GCMs sufficiently high resolution to allow extraction of this sort of quantity on the relevant scale? And what is the relationship between temperature and the different phases of hurricane life cycle? - presumably some sort of event analogous to crystal seeding (formation of a stable storm cell) followed by growth - and IIRC it was the intensity of maximum storms that was increasing, rather than their frequency (but remember Epsilon!). How does this play out in a hotter world, exactly?

PuD