The reason Lynas's book has re-calibrated my thinking is that so much of the current conversation of about climate change--and indeed much of the research on ecological impact--are outcomes in chapters 1, 2, and 3. Degrees 4, 5, and 6 have much less research to draw upon and we have to look back to long-ago eras to get a sense of what the world might be like under these scenarios. For example, the last time the world was 3 degrees warmer was ~3 million years ago during the Pliocene and there was no permanent ice on the poles and probably no glaciers. (Where will people get water that now depend on glacier melt?!) To imagine 6 degrees C, we have to go back even further in time, hundreds of millions of years before the present.
Lynas makes a compelling case that to avoid a climate that is profoundly different than the one to which we and most species are adapted, society should limit climate change to 3 degrees C or less. To make sure that we do not pass 3 degrees, we probably need to prevent CO2 concentration in the atmosphere from crossing 450 ppm. Right now we are at 390--so we do not have much room left to grow. If we move beyond 3 degrees, than the strategies humans might use to just "live with" climate change are probably not going to work--they simply aren't up to the task. (This is the basis of the "cruel hoax" that Romm uses when describing the prospect of adaptation without mitigation.)
One clarifying point is that Lynas's view of climate change takes a long perspective on the changing climate, and some changes that he discusses will take hundreds or thousands of years to occur. But they are a likely eventuality of changes that we make to the atmosphere today. For example, sea level rise will probably occur slowly, though the rapid ice melting we have seen recently implies that the rise could be faster than we previously thought. The same goes for the complete elimination of glaciers and for warming the Arctic so completely that carbon stored in the permafrost is released to the atmosphere. Lynas makes a very compelling case, however, that the feedbacks in the climate system that will alter our planet so strongly--even the slow ones--are things that we will not be able to control once they get started. And these run-away processes are further reasons to get our climate act together.