Re: So-
The process science is intended to follow is to form a hypothesis, test its predictions against reality and, if it fails to match up, recognise that and then either discard it for a new one or modify it.
A model is a complex hypothesis, nothing more, nothing less.
With "climatic" models we have three problems. One is that the timescales that get discussed are really too short to be called climate but we don't have anything in our vocabulary to deal with the gap between weather and climate. The second is that over those short timescales meaningful measurement is difficult at best and arguably impossible. The third is that both sides are so dug into there positions that they would be reluctant to recognise any detectable* failure of prediction.
*Personally I'm not convinced we're seeing anything but noise.