Also it's interesting that in that particular debate, there are "deniers" but no "zealots". It's only one side that feels the need to label the others. That ought to set the alarm bells ringing.
In fact, doesn't a denier automatically, by his mere existence, claim that "the others" are "zealots"?
True enough. However coming to a discussion with no opinion is next ot impossible for someone to do. It's reasonable to imagine someone who follows the evidence will have a pre-conception of what the answer is.
What sets him/her apart is a willingness to be proven wrong (or even just a bit less right).
If you already know the answer to some complex issue that is not direct, hard science, then you are cognitively limiting yourself. If it's reproducible and demonstrable, of course, then we can still remain properly agnostic and rely on induction to save the day. If it is not, we have to be very, very careful about what our hidden assumptions are.
I'd add one piece -- it goes both ways. On the other side of the skeptic is the true believer, which knows that the current consensus or opinion is correct and is just looking for reasons to confirm his belief. This is the scientific malpractice the leaked emails reveal.
Both of these positions suffer from the same problem: they already know the answer before the conversation (or experiments) even begins.
By the way, these topics go way beyond climate science, or even politics. These are the same types of conversations we have in technology teams about things like TDD, or linux versus windows. There are lots of opinions and consensus (or not) but most times what we really should do is have an open mind, create a theory, perform an experiment, see what works, and adapt. Many times teams would rather argue than do this. This is because we (technical people) suck many times at being able to politely disagree.
The same problems we see with skeptics and true believers in climate science are the ones we see all around us everyday. To me understanding and being able to work these kinds of conversations have much more immediate and useful value than the types of compression algorithms Google uses on its databases.