The testing numbers are not really comparable as different countries are using and reporting stats differently. For example in the UK they now only test those admitted to hospital, but there are thousands more managing the illness at home and "presumed positive" as logged by the NHS.
I'm afraid the death rate is the truest indication of what's happening, and even then I doubt every nation is reporting in the exact same way. It's really difficult to make comparisons and will be very interesting in months to come once all the data is available to see which methods and treatments worked best. Hopefully we'll end up with a decent global blueprint to work from for when a virus like this happens again.
There are so many problems with death rates (but you're still the best measure - although those dead people have to have been lucky enough to get a test before or after death).
First, in NYC, the fire department reports a huge increase in people found dead at home. Suddenly. Unusual. And EMT's and firefighters are testing positive. Ordinary autopsy procedures are suspended.
200 deaths more per day reported by NYFD. I trust their sense of (that's the last couple of days - probably the peak/near peak).
All over Italy, people dying at home and in hospital without tests and without autopsy.
Some nations not testing much at all (Russia, much of Eastern Europe, Saudi Arabia).
Younger people dying outside of hospital but not autopsied (several states, anecdotal).
SO, epidemiologists will use many measures (coffins ordered, cremations done) to compare this year's death rates to all the years where there is data. They'll look at incidences of known cases.
Some percentage will be attributed to the flu, using the best flu data.
But the predictions are that twice (thrice) as many people are dying in February, March and April than is usual.
I think it's 1.5 to 2X higher than what we see via the test kits. Maybe even worse.