>It would be impossible without a lot of guessing and estimations, producing questionable data, but you would have to include CO2 attributable to exports and imports.

>Which is just too hard, and too open to change assumptions to fit a desired result.

See: https://news.ycombinator.com/item?id=45762344

No, it's pretty straightforward. Count where a given good is consumed rather then where it's produced. It has to be estimated, but that's also the case for territorial emissions or other economic figures like GDP, but we don't throw our hands up and say "well it's too hard and too prone to fudging so we might as well not bother".

>Because in reality, much of the globe's economy is waaayyyyy too interconnected, and the arrows don't just point one way. Feedback loops without end.

What "feedback loops" are you talking about?

>That whole "this/that country..." just does not work, except to fill comment sections. The systems are global.

Ok but surely you must recognize that the US, where the average person drives a pickup/SUV to work is emitting more carbon than something like India where the average person gets around by walking or using motorbikes? That's the concept that conversations like "US emits more carbon per capita" are trying to capture. "The systems are global" sounds like an excuse to continue driving a F-150 to work because of some spurious arguments about how hard it's do to do carbon accounting 100% accurately.