Here's the 2007 IPCC estimate:
"For the next two decades, a warming of about 0.2°C per decade is projected for a range of SRES emission scenarios". [1]
Now, it did appear that estimate was slightly conservative. The actual warming for the first decade according to NASA [2] was about 0.27°C.
[1]https://www.ipcc.ch/publications_and_data/ar4/wg1/en/spmsspm...
But guess what. My second link, the NASA one, shows the global increase of temperature between 2001 and 2011 to be 0.03°C. The IPCC estimate was not that conservative, was it? Should I say it was aggressive by a factor of 6?
Nope, I'm not saying that. What I'm saying (and what IPCC is saying) is that generally their predictions appear to be in line with the outcome, and not to be very conservative.
Are we looking at the same graph? Because the graph clearly shows that yearly variation of ~0.2°C is quite normal, and picking any two years separated by 10 years is basically asking for random noise.
When IPCC said global temperature would increase "between about 0.15°C and 0.3°C" in 15 years, they didn't mean "between 1990 and 2005, every successive year will be between 0.01°C and 0.02°C warmer than the previous year." Nor did they mean the year 1990 and 2005 will be "average" years whose temperatures would be smack in the middle of the surrounding time period. To think otherwise shows a fundamental misunderstanding.
Climate scientists are not lawyers. They usually write for people with some scientific training, who can and are willing to follow their arguments for what they are.
Also, ".27 C" is not necessarily excluded from "about .2 C": it's not sufficient to simply compare approximate figures. There's a big difference in this context between ".2 C ± .5 C" and ".2 C ± .8 C". You would need to show that the observed result is entirely outside the predicted range, including both sets of error bars.
However, focusing on the temperature record is pretty disingenuous. It's been proven beyond doubt that humanity is liberating gigatonnes of fossil carbon, which is proven beyond doubt to warm the planet, specifically by 3.7 W/m^2 per doubling. The actual rate of warming and total climate sensitivity are interesting areas of research, but do not themselves have any bearing on whether that warming will happen.
The new updates are subject to the exact same bias though, it's not like you should expect it to start biased and then just immediately track towards "honest" results. Think of it as slowly reducing the bias over time, not pegging a starting point and then immediately jumping to "honest" where you expect the changes to fluctuate in a normal-distribution sort of way.