Thanks for the link. The effect is small at the 18oC avg surface temp. The slope of the equilibrium constant for carbonic acid, as a function of temp is the heat of formation/R(the gas constant). The solubility of the CO2 is depedant on the absolute temp, so a 1o change is only worth ~0.35%. That's a 0.35% change in both cases. On the other hand, the amount consumed by photosynthesis increases by ~10% for a 1o change. The reaction rate there ~doubles for every 10oC change. The net effect is a decrease in atmospheric CO2 for an increase in temp around the 18osurface temp average.
BTW, here's the article I was referring to earlier. http://www.rocketscientistsjournal.com/2006/10/co2_acquittal.html. If you look, you'll see that CO2 solubility changes can explain the changes quite well. But obviously temperature dependent increases in photosynthesis can't because the observed correlation is defintiely positive and quite high.
One of the things the guy notes is that in the period considered CO2 changes are time lagged relative to temp changes. That's part of his argument about why CO2 doesn't cause the observed warming. It doesn't address directly the question of whether today's human driven increases might.