That is why I added this to the discussion.
The objection is that if there is a continuing source of heavy gas, the heavy gas will stratify initially, because of its continuing supply, but after the supply ends, the diffusion will prevail and the stratification will end.
This video appears to show the opposite.
The CO2 comes from nearby lakes where some geological phenomenon shakes the CO2 and liberates it from the water. This is called from what I learn "limnic eruption"
https://en.wikipedia.org/wiki/Limnic_eruption
The CO2 doesn't originate where you see the CO2 polls forming, but elsewhere in the region on a lake.
The CO2 will accumulate in those low sacks of terrain but it is not formed there. If this is the case, that negates the theory that diffusion always prevails. The air of the "region" has an anomalous content of CO2, which is slowly collected into those depressions where it creates dangerous sacks. The polls are open to air, yet the CO2 remains concentrated in the polls.
The phenomenon is called in local language "evil wind"
https://en.wikipedia.org/wiki/Mazuku
The CO2 is not coming from the ground, but can accumulate coming from elsewhere.
Even if, as you might maybe think, the creation of CO2 from the limnic eruption were continuous (which I understand is not), the fact that the pockets are created far from the lakes show that the stratification happens regardless of the laws of diffusion of ideal gases. The lakes "burp" the CO2 and that accumulates at kilometres from there, in depressed spots.
The name itself, "evel winds", shows that the phenomenon is not continuous or stable, as it would be if the CO2 creation came from under the ground. In that case, the precise spot would be considered simply unsafe. But what is unsafe is the "wind", which carries the CO2 from elsewhere.