The paper in question [1].

Their big results seems to be that on a log-log scale CO2 emissions are linear with respect to population with a slope of 1.12.

[1] https://www.pnas.org/doi/full/10.1073/pnas.2501224122