Note the numbers used in the article are bogus, at least for Switzerland, where the number shown is closer to the number of tests per day, not to the total number of tests carried until now, which is 50-100 times higher, even at the date the article was published. Here in Switzerland we do about 2000 tests a day, for a population of 8-9 mios. Also we started testing relatively early thanks to the rapid availability of the Roche RT-PCR test.
This high number of tests may look impressive at first sight, probably the world highest per resident, however since last Friday the number of people with symptoms exceed by far our capacity not to test but to collect samples. The challenge is not about the analytical capacity (as in other countries) but about the collection capacity because collection by qualified personnel takes time with important safety precautions, not saying some pressure on the availability of nasal swabs. The reason why I'm now recommending the switch to at home self blood tests, this at least to know when people under quarantine should be discharged. China is again doing right with now using the IgG and IgM antibodies test as the main tool to discharge patients regardless if they tested positive or not to the RT-PCR test.
This does not change though the overall message of the article that yes the mortality rate numbers do depend on the testing strategy and that current estimate by WHO at 3.4% is most likely too high given the high number of asymptomatic subjects.