As far as I am aware of, carbonic acid may form when CO2 dissolves in liquid water. Most of the CO2 molecules in water exist as hydrates, only a fraction react with water to form carbonic acid.
In the reaction CO2(aq)+ H2O(l) <=> H2CO3(aq), the equilibrium strongly favors the left-hand reactants in such a manner that in saturated, ~0.033 M, aqueous (solvated) CO2, the ratio [CO2(aq)]/[CO3H2(aq)] is about 600:1. Heating reverses the reaction.
BTW, it is widely recognized that as result of this reaction, the sea contains about 60 times more CO2 than the atmosphere.
When it comes to the ability to neutralize base, all dissolved CO2, not only carbonic acid, is available because CO2 can react directly with aqueous hydroxide ion.
In short, it is my belief that CO2 is not supposed to affect water's dew point in combustion gases at atmospheric pressure, by forming carbonic acid as postulated by Roaling.