Also, it's important to realize that a lot of our data is "old". The tests for "cases" were probably queued for many days, and even when there is a result, might take a day or two to show up in the official stats. Beyond that, the infected individual has probably already been infectious for a week or two (or more?) prior to the test.

Put all of that together, and a lot of our stats are stuck weeks in the past. Since the phenomenon itself is exponential and apparently has a fairly short doubling time (two to five days maybe), that delay is huge. We could be flying into the side of a mountain, and we probably wouldn't even know it.

Given that, it'd be difficult to "overreact".