Cloud formation has been SHOUTED as a major weakness in the climate models forever. That is not the only one. Nearly all the elements of the GCMs are very far from capability of modeling within 1% their own contribution. Even 1% leads to a chaotic effect that cascades without any assurance the model is close to predictive even 10 years out, to say nothing about 100 years out.
and apparently they don’t take into account things like rainfall either