Here is a WHO report from February saying the same thing:
> Modeling is a helpful tool to try to account for missed cases, such as those that are mild cases potentially missed in current surveillance activities, and the time lag between onset and death. Using an estimated number of total infections, the Infection Fatality Ratio can be calculated. This represents the fraction of all infections (both diagnosed and undiagnosed) that result in death. Based on these available analyses, current IFR estimates range from 0.3% to 1%. Without population-based serologic studies, it is not yet possible to know what proportion of the population has been infected with COVID-19.
> Modeling is a helpful tool to try to account for missed cases, such as those that are mild cases potentially missed in current surveillance activities, and the time lag between onset and death. Using an estimated number of total infections, the Infection Fatality Ratio can be calculated. This represents the fraction of all infections (both diagnosed and undiagnosed) that result in death. Based on these available analyses, current IFR estimates range from 0.3% to 1%. Without population-based serologic studies, it is not yet possible to know what proportion of the population has been infected with COVID-19.
https://www.who.int/docs/default-source/coronaviruse/situati...