The great irony is that the stats have been criticised for being both too high and too low - people who get hit by a bus a fortnight after testing positive for covid count as 'dying with covid' or people who die 30 days after a positive test are missed off the stats etc etc.
But the fact is that there's always going to be variations and the stats are never going to be 100% accurate because there's so many different ways of counting as illustrated by the variations across different countries in the world both in the amount of cases and the percentage of serious illnesses and deaths amongst cases.
Is that country with a very low rate per 100 000 really free of covid, or are they just not testing?
How are deaths recorded? About a year ago, Russia appeared to have hardly any deaths from covid, but they had a lot of pneumonia deaths. It is very likely that a lot of these were actually covid. I recently read that the death toll in Peru has just doubled because they've either caught up with analysis, or it's being revised.
Is that country with the highest percentage of deaths per cases really much worse at treating people who are ill, or are they just recording stats differently.
Etc etc
The best estimates are going to be long after the fact in terms of excess deaths and in the meantime, we can't consider the numbers to be more than an indication, as there are endless caveats.