How can these numbers be compared when we really don’t know how many folks were able to be tested “when” or “if” they had the virus? In addition, the rate of testing varied between countries and here in the United States, we were quite late to the game in this regard. Wouldn’t all of this factor in to defining the results?
Your point is well taken, nut I think the delays in testing are denoted in the time constant to the first inflection point. Once testing gets going it appears to be fairly consistent. So it would make sense that it would produce a discernible time constant. However, this is indeed an assumption of this analysis.