On aspect that I find most interesting is how clear it becomes that a fact does not necessarily mean what one would want it to mean. For example, one of Will's claims is that "Antarctica is getting colder and its ice is getting thicker. RealClimate responds as follows.
…Thickening ice in Antarctica has been predicted by climate scientists for a long time, as a consequence of the greater moisture-carrying capacity of warmer air, so evidence for a thickening ice sheet would actually support, not negate, other evidence for global warming. …So even if we assume that the ice in Antarctica is getting thicker overall, that supports the underlying model of global warming that Will and Crighton are arguing against. Do Will and Crighton have an even more sophisticated model that captures the same phenomena but that predicts no global warming? If so that would be worth examining.
But the lesson here is that a particular fact does not necessarily support the more general conclusion that one supposes it might. A fact may provide evidence for a model. But it's the model that at least claims to capture the underlying nature of reality and that may allow one to make predictions.
For an earlier example of how Fox News used the same technique to mislead its viewers, see this post.