After The Economist, Politico has a stab at the 'International Morbidity League Tables', using the 'excess deaths' method, because:

How to measure the real toll of the novel coronavirus has been a persistent question since the outbreak began.

There are a number of difficulties. Some countries test more per capita while others test less. And governments collect their statistics differently, making any cross-border comparisons -- be it infections, recoveries, or deaths -- inexact.

Sometimes, a country does change its numbers reporting criteria: France started adding deaths in nursing homes to the ones in hospitals from April 4. The UK only started doing the same (only for England & Wales, apparently) last week, and both the COVID-19 cases and fatalities numbers suddenly jumped ahead of France (both countries have about the same population, around 66 millions).

A major rationale for this approach is that there's a lot of confusion around counting COVID-19 deaths, with different countries using different methodologies. Some apply broad definitions that include all possible cases, while others use narrower standards. That latter approach, however, has sometimes led to accusations that deaths are under-counted to downplay the crisis.

By contrast excess deaths can, in theory, cut through these uncertainties by providing a reliable figure.