The sky dragon response though is not helpful either. There is a radiant impact that varies with conditions. ASHREA did a cursory test for including a radiant barrier on the R-value of construction insulation. Under ideal conditions, a radiant barrier increased the R-value of a 3/4″ air space by 50% from R=2 to R=3. The problem though was that dust, condensation and orientation impacted the values, mainly negatively. That is pretty much the same thing that is happening in the atmosphere. There is an optimum radiant impact condition that nearly everything impacts negatively. Denying any radiant impact is lunacy.
A better way of showing the impact of “back radiation” is to separate the air temperature by individual thermal impacts. Kiehl and Trenberth’s cartoon was wrong on so many levels, but combining all atmospheric energy into one “Down Welling Long Wave” arrow was the worst. Had they separated the huge arrow into it individual components, like the NASA budget, they might have better illustrated the effect and found their 20Wm-2 error to boot 