What's the positive cases per test rate? I'm assuming that is going up, which is bad.

The one admittedly lagging indicator that I'm watching is hospitalized cases.

So far that's ok, and ideally you could possibly have an increase in positive cases throughout a population without more hospitalizations if high risk people isolate themselves. But yeah, hope it doesn't go sideways.

The site I posted indicates that there are almost 20 tests in the US for every new case, while there was only one test for every three cases in Brazil. Probably the latter ratio is not very representative without averaging over a longer period of time. However, it seems clear from the total test counts that Brazil has had more than an order of magnitude fewer tests per capita than in the US, and I am reasonably certain that the numbers indicate they are overwhelmed and the reported cases are the ones that they cannot attribute to anything else, so probably undercounted.

Here's something to think about - total cases for the world were up about 2% yesterday. US cases were up a little over 20K. If US cases were increasing at the average rate of the whole world, it would have been closer to 40K. Similar story in Europe, but more so; there more than half of the new cases are in Russia. So considering Western Europe, the epidemic has been greatly reduced and is growing elsewhere.

What's happening very quickly is that Asia and South America are now having a disproportionate amount of the new cases, even with China, Japan and S. Korea reporting almost none. And the spread is happening in countries where the testing capacity is relatively low.

If the lockdowns had had no effect, I can't see how to explain the dramatic shift over the last month in where the new cases are.

The general reaction of authorities to the epidemic seems to me very similar to an immune reaction. You have a strong initial reaction to an infection, and then it moderates. Earlier is better, but either too much or too little can kill you. Once it is slowed, you can go back to a relatively normal metabolism.