It is not only that countries are not accurately reporting these metrics, it is also that maybe they don't know. For example, if you don't know how many people have been infected, how can you determine a case fatality rate? Testing would certainly improve this but tests are not readily available and only people who are symptomatic are being tested. Does this mean that we should test everyone? Big problem here is the more tests you do, the less accurate the test is and the more false positives you will get (Bayesian inference; we have known about this since the 1700's and we are still not paying attention). Tests also should have been in the past, two weeks before any other data is determined. We don't have time machines. China has done a lot more screening than the United States. Mortality rates are also a problem. How do we know that someone died because of COVID-19 unless we test them?