On one hand, we have theory. There's reasonably good theoretical reasons to expect that increasing carbon dioxide should increase global temperatures. However, every good scientist knows not to trust theory too much until it's been verified by experiment, especially when the system in question is so complex.
On the other hand, we have experiment. Unfortunately, we only have a sample size of one, and we don't have a control group. We tried increasing the CO2 emissions, and found that sure enough, the temperature did wind up increasing. But given there's only one planet in the sample and no planets in the control group, we certainly can't eliminate the possibility that there's no causal link. If we forget that we're looking at global temperature vs CO2 concentration and pretend we're looking at a high school kid's science project about whether playing music to plants makes them grow faster, we'd have to say that the experiment lends some support to the hypothesis but is hardly conclusive.
So what can we reasonably conclude? I think the fairest thing we can say is that it's more likely than not that increasing CO2 emissions causes an increase in global temperatures. Anyone who claims the science is "settled" though, the way it's settled for heliocentrism, evolution or relativity, is not much of a scientist.
(I am also, despite years of Year 9 science projects, still unsure whether playing music to plants makes them grow faster.)