Richard Lindzen's Iris Hypothesis is being discussed anew with most of the same issues still in place. I am not really a fan of the Iris Hypothesis because it is more radiant energy related than plain vanilla thermodynamics. Most of the radiant models require quite a few thermo and fluid dynamic assumptions that I just cannot accept.
In the tropical ocean you have surface air that stays close to saturation most of the time. For a given temperature you have a saturation vapor pressure and as the temperature increases, the saturation vapor pressure increases and the dew point temperature also increases. You have more potential water vapor and a larger temperature range to ring that water vapor out. All things remaining equal, clouds should start forming at a lower altitude and persist longer. That should be a pretty simple negative feedback to increased SST.
Clouds forming at a higher temperature and lower altitude can create greater super saturation levels and produce more super cooled water in the clouds. Basically and increase in mixed phase clouds. Since the latent energy has to be released and higher temperatures/more CO2 would reduce the rate of release, thicker clouds at lower levels with more water/water vapor in super saturated or super cooled conditions would change the radiant properties of the clouds.
Lindzen's Iris just assumes that this will cause more efficient ringing or the moisture reducing water vapor entrainment to high altitude cirrus clouds. leads to a SWin versus LWout in clear tropical sky issue. While the lens lets more LW out i.e. you can "see" a warmer surface, the SW in can "see" the surface as well which would increase surface energy uptake. You are back to a square one radiant issue when the thermo indicates more interesting possibilities.
Part of those possibilities is that the mass of the atmosphere is pretty well fixed and just adding water vapor reduces the mass of a parcel of air, increased convection. That is only true for water vapor, once you get into supersaturated water vapor and super cooled water you have increasing mass of the parcel. There are regulating thermodynamic features included that aren't all that well considered in my opinion in the simple radiant models.
Since the mass of the atmosphere is effectively fixed, convection and advection would have to change with increased temperature. I liken it to a pot lid, more heat just makes the pot lid rattle more and that rattling is a bit random. The rattling, deep convection, is triggered at a temperature of around 27 C which in the tropics effectively limits the maximum average SST to about 30C degrees. There are somewhat isolated hotter pockets, often over shallower water, that can persist, but it appears to be unlikely that larger tropical areas can sustain greater than 30 C for very long. There are a number of "oscillations" MJO, QBO and ENSO that are generated by and help destroy these hotter pockets. So while the Iris Hypothesis is likely correct, all the mechanisms that would determine if it is a negative or positive feedback are not so easy to figure out.
More mixed phase clouds though is most likely a negative feedback and liquid layer topped clouds are definitely a negative feedback. This would indicated that tropical clouds are a regulating feedback pretty much like pre-CAGW science had them pegged. So until the complex mechanisms can be explained well enough, the Iris is likely to keep being debated. That is the problem with simple explanations, most of the time they aren't.