The thing about this is that it's a fluid situation and there is constant upward pressure on the death rates in these countries (the number of new deaths is a greater proportion of existing deaths than new cases is of existing cases). Germany was at 0.3% last week, at one stage South Korea was at 0.4, now it's at 1.5.
As of right now the best data for an entire country is without question Iceland and this really should give the clearest picture going forward. They've already tested 3.7% of the entire country. Above and beyond the government tests, they've also got a company (deCode Genetics) that's doing voluntary community testing. The government has a detailed and dedicated page for both:
https://www.covid.is/data
To put that in perspective, NY, which has by far the biggest testing program in the states had by yesterday tested 5,319 per million. Iceland has tested 37,371 per million. The current data out of Iceland is encouraging: 890 cases, 97 recovered, 18 hospitalised, 6 in ICU, 2 dead. That's obviously very hopeful but it's clearly not the end of the story either. For instance yesterday there were only 3 people in Iceland's ICUs. I think one of the things widespread testing does besides capturing more people with genuintely asymptotic and milder conditions is that it captures them at earlier stages the disease. There is therefore a greater lag between discovery and death in these countries. That needs to be factored in when attempting to ascertain the true death rate.