I don't doubt it. But they cannot be a dominant contributor globally unless the Vostok data are wrong. The fellow is saying that the oceans are the dominant sink, that their efficiency is temperature dependent and that variation accounts for the historical CO2 concentration record.
Sure [CO2] [is not insulated from the rest of the atmosphere]
No it isn't as you yourself note: they bang into other stuff and there's an energy transfer. Consequently heat will flow from the CO2 to the other atmospheric gases. By my calculation, the N2 and O2 in the atmosphere will take up about 2000 times the amount of heat as the CO2 given the relatively minute amount of CO2.
The Oceans are essentially saturated. I'd have to read more about Vostok ice cores to know exactly what they're about. Still, the change is small and is irrelevant for GW concerns.
"Consequently heat will flow from the CO2 to the other atmospheric gases. By my calculation, the N2 and O2 in the atmosphere will take up about 2000 times the amount of heat as the CO2 given the relatively minute amount of CO2."
I'm not sure you understand how this works. Polar molecules are the only absorbers. If there were none, only the surface stuff would absorb radiation. Since there are polar molecules in the atmosphere, it too absorbs. The amount of energy absorbed is proportional to the concentration of polar molecules. Even though CO2 is at a low concentration, it is responsible for a fixed definite amount of solar absorption. Doubling the concentration, doubles the value of it's term for the total energy absorption. The retention depends on the polar gases, because the other gases don't absorb and are fixed. Maybe this link and the calc and discussion with JasonC in #7 will help. Granted I don't know the details about Vostok yet, there temp numbers says the Sun's output varied ~10% over those cycles. So the sun's radiation varied from say, 1235W/M2 to 1365 W/M2. The apparent concentration of polar only changed +/-17%, that's too small to effect a significant temp change.