Is this difference due to some property of population (e.g. average age, smoking rates) or the response?
Additionally, it seems SK has picked up cases much earlier in their progression than Italy (and most other countries except perhaps Singapore). This means we'll likely see more deaths as those cases resolve (and deaths/cases in SK has been increasing, was 0.6% a week ago, now 0.9%).
Another factor is it seems SK picked up a majority of their cases, mild and serious, due to extensive testing, while Italy has likely not detected many mild cases.
Finally, death rates are much higher in places where health systems are overwhelmed, such as northern Italy.
You should estimate the levels based on the sampling. For example last week the UK Govt said they estimate there was 10,000 infected in the UK based (in part) the number of positive people in the hundreds. You should be able to make better predictions using geo-demographics and population density etc.
For south korea, do we know the number of people who are in hospital with the virus being treated? Then you can compare Italy and South Korea.
[ https://tg24.sky.it/salute-e-benessere/2020/03/16/morti-coro... ]