On the observational determination of climate sensitivity and its implications (full text PDF)I think it's a much more refreshing reading than the biased reactions at PNAS where the paper was previously submitted.
The authors have admitted several errors of their 2009 paper that were pointed out by subsequent papers and they believe that they have addressed them and the mistakes are fixed in the new 2011 paper.
The measured response of the energy fluxes at the top of the atmosphere to the changes of the sea surface temperature is still significantly larger than claimed by the IPCC and the climate sensitivity ends up being 0.7 °C. In fact, their 99-percent confidence interval is 0.5 - 1.3 °C. Imagine: they are 99 percent certain that the climate sensitivity can't exceed 1.3 °C. This is, of course, qualitatively incompatible with the IPCC whose lower bound is 2 °C.
Quite generally, I like the fact that their method uses lots of data point - a more detailed measurement of the responses to the atmospheres to many episodes of warming and cooling of the seas. This is the real source of their pretty impressive accuracy.
If the calculated sensitivity 0.7 °C were pretty much right, it would follow that one expects no warming by 2090 or so - when the CO2 will be twice the pre-industrial value - because a warming by 0.7 °C has already occurred so there's "nothing left", assuming for the sake of simplicity that all non-CO2 sources of warming average out over two centuries or so.
Feedback about the paper will be appreciated. It would be great if some TRF readers actually read the whole paper.
