Non linear systems are much more difficult for obvious reasons. It's much harder to predict out put if it isn't a straight line. Now add in chaotic systems. Now add in multiple feed back mechanisms not fully understood and possibly interpreted wholly wrong. That gets you close to climate modeling.
As already noted, the problem with cd's graph is that the PPM is expressed in a way to make it look bigger than it really is as a percentage...parts per hundred.
That graph shows an increase of .00005% in atmospheric content of Co2 since 1750. While the actual concentration of the gas has risen from .000272% to .000338%. Very little.
It's not generally expressed in PPM. It's 0.10% not 0.0000382% It's off by a factor of ten thousand.
Theorizing is one thing, empirical facts are another. At this point, both sides are theorizing.
The Climate scientits circular logic is weird. "You can't prove us wrong so we must be right but you must prove us wrong even though we can't prove we're right."