This is an answer you cannot get from observing a plant with human eyes and senses, so nobody can answre this in a specific way. It's about various rates of change interecting / contradicting each other etc. Large samples, precise measurements and repeated experiments needed.
Sure, you get less evaporation with cooler temps or higher humidity (vapor pressure vs condensation pressure is the 'gradient' mentioned. these are 2 rates that result in net loss/gain or equilibrium over time, plus in this context you also have rate of production, too), but it also will reduce the amount of potential photosynthesis per day - which reduces co2 use. Products of photosynthesis are needed to power molecul building of cells -- whether cell reproduction or just shit cells make to function and live.
If you do something that reduces potential photosynthesis, less CO2 will be used. You could maybe extrapolate based on photosynthesis vs temp curves... i don't know if it's 1:1 proportional, though, but that info too could easily be out there already calculated. If you reduce DLI 10% you can reduce CO2 X% etc... You will still benefit from excess co2 as long as the temp/rh part of that equation doesn't pull it back down to 'ambient levels' as a ceiling... the 'max' co2 you can use is a product of the climate variables. if those variables calculate to ~300-600ppm, your ambient co2 then becomes 'enough.' local measurement needed, of course.
now, at what point does focusing on limiting terpene loss cause greater net-loss due to a slower rate of terpene production or reduced maintenance (involved?) of terpenes in the plant? this is what i mean by numerous rates of changes being involved... this is calculus and to study any one component a time, you'd need to tightly control all the rest. then repeat experiemetns numerous times over to verifiy results / conclusions. It's a slow process. Then repeat with each relevant factor or combinations of factors to paint the full picture of reality.
90% of initial studies are later debunked, which shows how much work it takes to prove just a correlation. Simple contexts are different, but this is not one of those. So, if you find somerelevant research, keep that in mind. if it's a simple situation where an obvious measurement is taken and you read over the methods to make sure they are sane, you can have more confidence in that context.
if you focus on any one bit it seems like you can do something that will have an obvious positive effect, but it also impacts other parts of the overall equation, which can easily negate those benefits.