When you talk about the AR4's "devastating" projections are you taking into account the time frame?
Well, I'm sure you've heard of GlacierGate (2035, and all the glaciers in the himalayas were supposed to disappear). Yes, I know these errors have been already admitted, but they're sort of what IPCC AR4 ended up representing as highlights...which the sensationalist press took on as gospel truth, which when debunked got trumpeted from the rooftops by every right wing kook that thinks evolution is wrong, so on and so on.
When it comes to the rate of change, I'm not convinced that we're looking at accelerating changes from 2000 years to 200 years -> my bet is that climate sensitivity is actually low enough and filled with enough negative feedback effects that our contribution is negligible. Of course, I could be wrong, but from what I can gather of the poor proxies that we have to look at in the past, we've already seen historical abrupt changes before.
On top of that, I'm of the firm belief that if global warming was real, and was happening due to CO2 emissions of man, we should encourage it -> a warm world is a better world for humanity in general, and since even though we might see a dramatic shift in 200 years, the upper bounds for any positive feedback effect will stop it from being a runaway situation. If we could re-enter the medieval warm period tomorrow, and stay there for thousands of years, it would be beneficial to humanity, especially considering the poles end up doing most of the warming -> we're not talking about los angeles become 120C year round, we're talking about the upper latitudes going from 40C to 60C...further, I think there's a pretty good argument that our increased global average isn't because of higher maximums, but more from higher minimums (that is to say, our winter months are getting hotter, not our summer months).
Anyway, in 20 years, you can buy dinner, and I'll buy the drinks :) Thanks for the interesting conversation :)