In short, if they can't get it right for next week, what leads them to believe they have it right for the next 100 years?
I've heard it argued that long term climate forecasting is easier than short term weather forecasting. Of course, that has yet to be demonstrated.