Let me use an air conditioned/central heated home as an analogy to the planet.
We've got some thermometers scattered around the house and they read differently within a range. We know if we add some insulation the house will retain more heat in the winter or keep it out in the summer. The exact change in a thermometer on going from 3" to 6" insulation isn't known, but everyone would agree there would be a change. We can calculate what the result should be but we also know from practical experience that the actual results will differ.
co2 == insulation.
Our central AC/heating unit has output measured in BTUs or watts. The thermometers only give us
an indirect measure of the effectiveness of the "Insulation" on the BTUs.
Now consider the usefulness or lack of in proxies - tree rings, ice cores, etc. Scientists do their best to derive "temperature" from these proxies. So we have an estimate of past temperature (an indirect measure of heat capacity of the system) being compared with current thermometer readings (plus adjustments but let's not go down that road for now).
I'm thinking temperature is the wrong metric for global climate. Actual heat capacity in the system is the metric, right?
So why are scientists trying to get temperature proxies?