Being able to fine tune 14 different climate models to track historical records isn't the same as modeling future climate behavior.
Models are "tunable" which helps to ensure they adequately model known history. That doesn't mean they will adequately forecast the future. I don't believe your climate scientists adequately understand the complexities of our climate well enough to model it. If they did the IPCC wouldn't need to use an ensemble of many models to try to forecast future warming -- they could just use the one that did it right.
I think it's very illuminating that you chose to use a graph the IPCC used to answer a question in an FAQ do ent but, wasn't actually a part of any IPCC report. It would be in their interest to come up with a pretty impressive example, wouldn't it?
Why don't you show me a model for surface temperatures produced in 1970, predicting surface temperatures over the next 40 years and then, lay actual observations over the top of it. That'd be more convincing than showing me a graph where the IPCC was able to tune their models to kind of, sort of, agree with observation.
If 90% of the doctors I consulted recommended that I undergo $500,000 worth of treatment in order to potentially prevent some disease that a computer model predicted I may get in 30 years I'd probably pass on their recommendation even though they have a consensus.