The fatality rate of countries with active epidemics is skewed by a variety of factors:
- With exponential growth, most cases will new and so not become potentially lethal. This can lower the observed fatality rate (I think this explained Germany's original "great" rate which doesn't look at great any more).
- With cases rising quickly, most countries don't have infrastructure or the time to increase testing (US testing is failing on multiple levels but even France, Spain and Italy are just overwhelmed with the sick and testing is less important).
- With cases rising quickly, the fatality rate increases as hospitals are no longer able to provide adequate care.
- Different countries have different age-profiles and the disease hits the elderly harder (but once hospitals break down, the odds for the young decrease).
South Korea's fatality rate is similar to the ostensible fatality rate of Turkey and Luxembourg but likely for different rate. We probably won't have a complete idea what's happened until these events are done.