I think the author oversimplifies the problems with climate models. They've had numerous mistakes and extremely critical ones. Not least of which is that they haven't factored for Multi-decadal changes in cloud cover albedo. The climate system is incredibly complex and our models do not have good models for all the subsystems that make it up, including the oceanic oscillations like Atlantic Multi-decadal Oscillation, PDO, etc. We can't even predict many of these subsystems with any degree of accuracy so we are hopelessly inaccurate at the higher level.