^^^^^hmm.......... interesting, but the curious dynamic in the graph begs the question - why?
Has the virus now mutated to make it less deadly ( but more virulent), have medical practices become more informed (and therefore more effective), or are deaths in the US now being allocated differently when there are other underlying other issues that coincide with covid, and death occurs? Or a combination of these?
Probably a combination.
I wouldn't necessarily say "mutated", The most successful viruses are ones that don't kill their host, and in fact, don't make the hosts so sick that they self-isolate, but that spread with the greatest ease. So some natural selection toward less severe and more virulent is to be expected.
Medical practices have indeed improved. We no longer blow people's lungs out by putting them on ventilators and cranking up the pressure. We've figured out that for bad cases, certain steroids can be quite helpful (which implies that it's not the virus itself that's killing people, but rather an immune system over-reaction to it).
We're also discovering (and thus reporting) a lot more cases than we were a couple of months ago. Hospitals are open for non-emergency procedures again, and they're testing everyone who's being admitted. Go to the hospital for surgery of some sort, get tested, have the test come up positive, and bam, you're a reported Covid case, whether you ever have any Covid symptoms or not. Early on, the only way to get tested was if you had severe Covid symptoms to begin with. Now, pretty much anyone who wants a test can get a test. Got a sore throat? A headache? A cough? Get a test. It's positive? You're a reported Covid case, whether the symptoms ever get any worse or not.
Then there's the demographic aspect. The people who are out and about and catching it are the young ones. In Florida, the peak of the age distribution curve of all these new cases in in the low twenties. These "kids" are highly unlikely to die of this virus.
I do not think that cause-of-death coding/reporting is a significant factor. If you look at all-cause mortality, it was clearly above seasonal averages for a while this spring, but has now reverted back to normal levels.