Suppose there's a 10% chance that the Gulf Stream will turn off (an event which has previously occurred, and which was followed by extreme rapid cooling of northern Europe), given one set of policy choices, and a .1% percent chance, given another set. How much should we be willing to pay to implement the second set?
The thing that always gets me about this debate, and the energy debates in general, is that energy efficiency is still far and away the most productive energy investment around: what is needed is ways for individual consumers to make those investment easily.
That logic doesn't work. You might as well argue that traffic laws are inappropriate because you almost certainly won't get killed on the way home today.
This is called "risk assessment", and it's an important and crucial part of any policy discussion. You don't just pick the average and assume that will happen, you weigh costs vs. risks across the board, including the "best" and "worst" scenarios, and try to come up with a safe, conservative plan for avoiding the worst effects.
I think you're misunderstanding the use of the "worst case" in the IPCC report. Those aren't OMGWTFBBQ scenarios from someone's dreams, they are the scenarios from the actual science that lie within a reasonably tight bound. They are likely consequences for some value of "likely" (probably a 95% confidence, or the like).
