What you say is likely but I would also point out that the most significant covariate with the death rate was age. So if you compared any two populations and if one had a higher average age, they would have a higher excess death rate. So any additional inference from another covariate should be done in a Multivariable Regression where you include the most significant covariates alongside your proposed new covariate.

A point in your favor would be to include BMI or some other metric for obesity, as that was the next most significant covariate after age and fod US vs Japan would counter some of the age related impact on the statistics.

I know you didn't perform this research but these are the types of things to watch out for. I'm this manner dishonest researchers can get whatever outcomes they want.