Let's see if these fare any better than the projections used by the IPCC ... http://esr.ibiblio.org/?p=5297 "The most recent climate model simulations used in the AR5 indicate that the warming stagnation since 1998 is no longer consistent with model projections even at the 2% confidence level." Kudos to them for publishing these, though. If they are rotten, it ought to be easy to tell.
http://www.theguardian.com/environment/climate-consensus-97-...
That doesn't feel like a particularly strong response to ESR's claim. Taken as a response to that claim, it feels kind of like: "our predictions were wildly inaccurate, but it's a complicated system with short-term and long-term factors and the short-term factors caused more warming just before we made our predictions and less just after, and that threw us off".
Which doesn't inspire confidence. If you can't predict the short-term factors, you need to widen your confidence intervals. I feel like at the very least, you should be making predictions like "conditional on these factors staying within these bounds, we expect this amount of warming". Then if that condition doesn't hold, you don't lose any bayes points.
(I recognize that the article wasn't written as a response to ESR, I'm not even sure it was a response to the same thing that ESR is talking about. But you offered it as a response, so that's how I'm evaluating it.)