Climate models are correct indeed. Because, as Bill Clinton would love to say, it depends on what the definition of “correct” is.
In the real world, climate models cannot be falsified by a single observation (modelists say it’s “only weather”), or by a set of short-term observations (they call it “just a specific trajectory”).
In theory, one could wait a sufficient number of years in order to statistically check if the world has actually got warmer, but in practice models don’t include volcanoes, clouds, solar activity, etc: therefore, even if observations diverge from the models, all the modelists will do is find a “culprit” that can justify the discrepancy.
For the 1940-1970 cooling climate, they say “it was the aerosols”. Never mind that it could be a made-up story.
RealClimate’s own Gavin has said in the recent past, there is no interest in verifying if models are correct or not. Instead, the “right question” appears to be: “are there analyses that will be made over the next few years that will improve the evaluation of climate models?”
It should go without saying that in such a situations, models have no predictive capability beyond chance and they are for all intents and purposes useless.
Imagine modelling a human being as a heart pump with tubes coming in and out, and then when the patient dies of tuberculosis, having the superciliousness to state “the model is correct” instead of understanding that humans have a pulmonary system too (and a lot of other systems).