I keep going back and forth: one day thinking we’ve overreacted, the next day thinking I’m not scared enough.
That said, the flip side to all that you posted is that for every death confirmed from COVID-19, there may be a similar percentage of COVID-19 deaths not attributed to it. I know of a small rural community that I’m pretty convinced had at least 3 deaths in February that sounded an awful lot like COVID-19.
With most statistics, the Devil is in the details. They have tested 3300 people and are claiming that 2.5% to 4.2% were positive, anywhere from 83 to 139. They are then extrapolating that to the entire county and comparing that to confirmed cases. While I'll call that very encouraging, that is a long way from compelling evidence.
A few questions come mind:
Why the range of positive, were they positive or not?
What is the sensitivity and specificity of the test?
How random was the sample? Was there any methodology to selecting those tested?
Assuming the test is accurate, how do results from Santa Clara county apply to the rest of the country?
The mortality of the 1918 flu varied greatly by region, probably due to nutritional differences...0.7% Texas vs 3.14% Connecticut. Sunshine, perhaps? This test was done in LA, which is pretty far south and quite sunny. The results there may not translate well to other areas.
That does give some insight into relative risk by region though. It might turn out that some areas aren't at as great a risk of crisis, and this could make it easier to tailor the restart/future mitigation by regional risk. I'd be very much for that.