The creates an inflated number that was convenient for politicians/health orgs to drive whichever policy they desired. Data was usually massaged to justify pre decided policy, not the other way around (common occurrence in economics/politics and think tanks).
Also, "GDP per capita based on purchasing power parity (PPP). PPP GDP is gross domestic product converted to international dollars using purchasing power parity rates": that's also likely to trip people up.
It would seem that the most likely explanation is that richer countries are better at confirming deaths? Or that small, dense urban microstates did badly? (Hong Kong the massive outlier)
It might be interesting to replot these vs median age.
(I prefer log-log as none of the scatter gets smushed; what would you prefer?)
How can this then be "confirmed deaths" if they themselves state it's not possible to accurately determine them.