My apologies, I was wrapping up my time online and looked at the graphs in a rather cursory manner. The graph I looked at (12.3a) the vertical axis is in Watts/M2 not degrees C, huge difference, my mistake. Dismay at the IPCC remains, but is mostly focused on the conundrum about why they are increasing the level of conservatism of their estimations, particularly when the science has been steadily demonstrating the errors of that practice throughout AR1-4. Findings that increasingly support higher CO2 emissions than earlier studies estimated, over shorter time frames, and generating stronger environmental responses from such factors. I hope the IPCC reviewers are more correct this time than they have been in the past, though reducing estimates doesn't seem to be the right direction to go when you have consistently been under-estimating effects more and more with each report released thus far.
To the extent of my knowledge, IPCC panels are a little bit conservative as they should be. And there's a giant misconception about the number and validity of those papers supporting "it's worse than we thought!". If I have to describe at a glance the difference between my involvement in these matters at forums.randi.org (and other fora) during 5 years as opposed to my involvement with these matters in real life for a longer time, I would say this:
a) Internet debate is clearly denialist-driven and pretty much unscientific.
b) Internet debate has little connexion with general education of the big public. It's like here; they can't teach what a strawman is because themselves ignore what it is, that's why so many strawmen are stuffed here and pass undetected while the same posters accuse others of strawmanization, for instance, to censor their ironic remarks.
c) Internet debate, as part of its unscientificity -four-dollar word if any-, is pretty much innumerate, asystemic and barely logical.
d) Internet debate is uncooperative, and elements of life that are completely alien to science take control, like proof within the adversarial system of justice, propaganda, political advocacy, etc.
Within this depiction falls a sad fact involving scientific papers, popular science and all that lies in between. The fact that a paper from 2009 finds some scenario laid out in 1995 to be wrong or improvable by using models, leftovers (like the microns in that sad sea ice debate) and by doing so it tells everything is worse than expected should be a warning about the quality of such papers. Late papers are most probable to fall into the opportunistic-to-make-a-name-for-ourselves category and not in the breakthrough one. On the contrary, and what a surprise in that, there are much fewer late papers saying that everything is about the same, with more accent in this and less accent in that, and even fewer papers saying it's better than thought. And nobody is citing them here (or there), why? because they don't satisfy the wished level of confirmation or alarm.
There's no worse and abject cherrypicking than googling and selecting a trio of abstracts from peer reviewed papers with similar conclusions. The right thing to do is selecting three because of their inherent quality, from a group of 10 or more, including the willful search for opposite conclusions.
Denialists had the public perceiving that AGW yes and no has a 60% - 40% share in the scientific production, so the question is still unsettled. In reality it's 97% vs 3%, the question is settled, but warmists have the public perceiving that those 3% are bad papers and those 97% are mostly good, what is as shameful as what denialists do, specially when they cherrypick from that 97% not for quality but for conclusion.
That's why IPCC ARs look so complicated yet a little bit hesitant. Science is there.