I think no one argues with what you can measure. The problem is with the claim that something extraordinary is happening. It's not clear how we can claim that, given the uncertainty produced by the models, the uncertainty introduced by making up 15% of data on input of the models, the uncertainty inherent to the stochastic process and the uncertainty of the proxy measurements for paleoclimate and other generalisations (eg even 100 years ago there was no direct measurement of the global temperature, only proxies and models).

Ok, we have a theory of human-made warming. What's the theory of Jurassic warming? It's got to be verifiable.

Ice cores, tree rings, etc aren't direct measurements. So there is some model to go with it. How do they define confidence intervals for these models?