COVID19 lethality correlates with intensity of exposure. Severity of government and individual countermeasures correlates with hospitalizations. There's a delay between exposure and hospitalization.
So the typical pattern is that individual and government countermeasures undershoot, leading to an hospitalization spike, leading to overshoot.
Vulnerability to COVID19 varies by country, culture, weather, population density, demographic, etc.
The USA certainly fits the undershoot overshoot pattern. But comparing the USA to a country such as Sweden, and concluding that lockdown was unnecessary for the USA, is apples to oranges. The population of the USA is very different than Sweden's.
There's no fixed number for COVID19 CFR, because intensity of exposure matters, and that intensity is self-moderating due to human countermeasures.
A purely selfish authoritative epidemiologist should predict a death rate that takes into account the effect of his advice on the death rate, which requires understanding not only the contagion but also societal preferences and competence.
It's much like making public predictions about the stock market.
Authoritarian China can tolerate higher casualty rates than Western democracies, which is one of the reasons the real CFR declined as COVID19 went global.