One item about the CO2 "model" is important, I think, as one can calculate rather easily the impact of more CO2 ALONE as a single variable on global temperatures, and the effect really is quite small, a couple tenths as I recall. Here is where it gets tricky of course.
Another issue is that a SLIGHT imbalance in the heat in/heat out ratio can over time exert a large change in global T. Heat is not the same as temperature of course, heat can be a far more influential factor. Imagine the heat balance shifts by 0.1%, over time that can be crucial.