Yak, no one knows the total answer. Beretta is right, nature has a massive lag effect. Oceans and vegetations store heat and only release when equilibriums they are involved in lose heat (i.e. cooling, yes nature knows what it is doing, surprise!) In times of excess heat, nature absorbs it and times of cooling releases it.
GHG phenomenon, is more to do with the reflection and capture of the radiation that is released from the surface. Obviously higher the concentration of particular gases the more likely that it will trap the radiation (its more complex than this as gases can only "get excited" by particular frequencies.) H20 has the biggest spectral range and account for 60-70% of the GHGs (another surprise that nature keeps the master key close to its heart!)
The reason why we havent seen the predictions eventuates boil down to a number of things which I dont think anyone can argue:
1. Everyday there is new theory/new natural phenomenon found that has previously not been factored. 2. Most of our predictions are model driven. We are modelling random complex interrelated processes. I can tell you we cant even model some exotic options properly haha. Its quite funny to me when I see climate scientists claiming to be modelling such processes, obviously the finance industry is not really keen on money. 3. The basics are correct, GHG science, human emissions but because of our lack of knowledge and the urge to get something published - yes it exists and the bigger the scare the more brownie points you get, we tend to have over sensitive model output. 4. The question is, even if there is an associated rise in temp to CO2 and correlation works out, does this prove a cause? We only have 1 sample period, we can not rule out false positives and statistically speaking it just loonie to say one observation proves theory. 5. Its no suprise and even submitted by IPCC that radiative forcing is a function of log of gas (Co2 if you want an example). Obviously since 70% of CO2 is non human, most of the log-effect is already accounted for even before we even emit anything. Note I am not even including the effects of enhanced photosynthesis and natures own equilibrium shifts to account for us as an error term which would easily handle some if not all of the 30% we emit.
I think the most important points are over-sensitive models, lack of respect for natural forces and log-effect.
