Well, instead of repeating myself manually, I'll paste in a comment of mine here from a past discussion on carbon capture:
It's easy to forget why there is a bit of a challenge to getting C02 out of the air: there's so little of it, comparatively.
In order, air is, broadly, made up of the following:
Nitrogen: %78.084
Oxygen: %20.946
Argon: %00.934
CO2: %00.042
The stuff is essentially beyond a rounding error - it really gives one an appreciation of the "either don't release it, or capture it at the point of release" sentiment, and for the difficulties in making carbon capture outside of these scenarios be even slightly cost-effective.
What if one started emitting Nitrogen, Oxygen and Argon in the right proportions instead to get the mix right again?
So then, is it really the CO2 that produces the cognitive impairment, or is the CO2 here just the proxy value that we are measuring, and the real reason for the cognitive impairment is low oxygen?