I get it, but consistency in presenting data is important when people are looking at it repeatedly to follow trends. The habit of presenting deaths first started when testing was much less available, but also because deaths are a much more reliable (though lagging) measure and the most important outcome. There are many variables posed between cases and resultant deaths, and some of those are changing over time, the other reason the CFR is changing (besides increased testing).
I think that's presenting the more interesting data first. Deaths should lag diagnoses. A spike in cases is interesting because it might predict a spike in deaths. Obviously there are a lot of other variables, but I would expect to see a spike in cases a few weeks before a spike in deaths. We've had about a week of a very high rate of new cases, so deaths might spike in a few weeks.
If this were a baseball game, the US would be winning by a run, but COVID would be at bat with the bases loaded with no outs. The interesting part of that game wouldn't be the runs that already scored, it would be the runs that might be scored in the near future.
Edited by porter, Today, 07:19 AM.