“Picking and/or choosing a specific date to represent a “trend”, tho, is deliberately misleading.” I fully agree with you on that point. Trends and totals are probably much better indicators. Totals and trends as percentage of population and percentage among those tested would probably be useful as well.

I’m choosing large swings to make a point with these examples and using the royal “you”.
If you have 0 deaths one day but 100,000 the next and then 500,000 the next the significance of your 0 death day is pretty much meaningless. But going from 100,000 to 70,000 to 30,000 to 10,000 to 5,000, to 1,500, to 750, to 300, to 2 would be starting from a pretty bad place but damn did you mitigate the hell out of that so good job on that front.

If there was a cheap, or free, way to test at home I think it might be kind of cool to track infection rates that way. Ask for voluntary compliance that everyone test themself every other day or some other determined interval and report the results. Give everyone that participates a random patient number that can’t be traced back to an individual may relief the privacy concerns people are bound to have. I’d like for it to be tracked down to at least the county level in the USA. Could use previous census numbers to estimate percentage of people taking and reporting the test. Also people that are asymptomatic might be smart and stay home until they test negative again.