Ciao Massimiliano,
this is a interesting solution. I’m curious about the result of your analysis, since I see a 3° discrepancy in a short range. Have you also taken into account the reflected radiation from the walls/context material? This seems to be a hot topic, seeing recent posts, and I believe is also the @ricardo goal.
I ask, while seems im moment not possible to reach this result, like this recent post Possibility to simulate in LBT the outdoor surface temperature caused by solar specular reflections - #8 by chris