Thanks, glad it was clear, and I will say that I had an advantage because I had written about this to a friend earlier yesterday so I already had most of the language worked out.
They had mentioned the worldometers data saying deaths were low and seemed to be leading the cases data. That struck me as odd so I took a look and sure enough that’s what the data showed. But I vaguely remembered that FL had changed their reporting method so I googled it and read a couple articles and ended up concluding that that would probably explain it. Then a little later when I saw this thread and your post I figured I’d say something since it was pretty fresh on my mind.
I agree a month or two is probably enough settling time for the data. Like you say it depends on paperwork processes playing out. If there were a way to “play back” the data for the last six months or so to see how it moved, you could really get a feel for the settling time, but that seems like a safe estimate.
BTW I used to do motion analysis using high-speed imagery...lots of looking at data curves in that world.
For me it was actually the steep rise in fatalities a bit earlier, but again leading the case data, and not by a small margin. That was clearly impossible unless:
1) A much more deadly variant was moving in, but was not yet dominating cases (but there were no other indications of that);
2) Fatalities data was being outright fabricated to attack DeSantis (tin foil theory: Clerical mischief to create spikes for headlines is quite doable if no one cries "foul", but major alterations of the 7 day average and curve shape in the manner observed require either massive numbers of fake death certificates or a change in reporting method);
3) The health (critical) care system in FL had collapsed (which it clearly had not);
4) Doctors were executing large numbers of Covid patients to attack DeSantis (extreme tin foil, but we have a contingent of FReepers who'd believe it in a heartbeat);
5) Indeed, a quirk in reporting method was in play.
Golly, I'd love to do a time lapse of some of these curves, but, I just don't consistently have the time, nor do I have a host to post them on.