Differences in the efficacy of climate forcings explained by atmospheric boundary layer depths by Richard Davy and Igor Esau (Nature Communications no. 7) explains a few things I have been harping on for a while. When you have low heat capacity you get bigger temperature response. Pretty simple really. There is a bit of a disconnect between the simplified GMST and change of actual energy in the system at the extremes of heat capacity. Well worth a few minutes of your time to peruse and worth more time if you are confused about cloud forcing/feedback.
I have tried explaining it with effective radiant energy and heat capacity to show how the zeroth law of thermodynamics rears its ugly head when you attempt to use an average temperature based a range from -80C to +50C combing "surface" air temperature with bulk ocean temperature which is about as huge a range of heat capacity you can find, but Davy and Esau have an approach that is much more likely to be acceptable in the climate science community.
Because most of the positive cloud long wave feedback is in low heat capacity situations and most of the cloud negative short wave feedback is in high heat capacity situations, it should be pretty obvious that overall cloud feedback is most likely negative if you are concern with increased warming in a thermo relevant way.
So if you have wondered about the planetary boundary layer aka atmospheric boundary layer impact on estimates of sensitivity, this part is a got start. Unfortunately, it doesn't delve into the issue of longer term impact on "global" heat capacity. When you have high latitude warming in winter that doesn't result in heat storage below the "surface", that warming is actually cooling if you consider Ts=lambaRF +dQ, because dQ can be negative. Ignoring the dQ just implies a higher sensitivity than actually exists. Since solar in the lower latitudes has the largest impact on heat capacity, it would have a higher forcing efficacy.