[BostonCard]

Last point:  I have to take this moment to re-raise the issue of overreporting of Covid deaths with BC, who was adamant that all prior reporting was accurate (and frankly, did not understand why I was asking if the deaths were "because of" Covid or died "with Covid").

That mischaracterizes what I said.  Of course not every single case could be perfectly adjudicated, but if a patient dies from a heart attack, the death certificate will list heart attack as the cause of death, even if the patient happened to have COVID at the time, unless there is reason to believe that the COVID caused the heart attack.  It is impossible to adjudicate every case perfectly, and there is going to be some variability and biases among physicians, but I think the it will all more or less balance out and broadly you can trust the numbers of deaths that have been attributed to COVID.

An easy way to reassure yourself that the numbers are broadly accurate is to look at "excess deaths".  If you count up all the deaths, regardless of cause, since the pandemic began, and compare it to how many deaths you might have expected based on what has been reported historical (plus trends) it turns out that the difference corresponds pretty closely to the number of deaths that have been reported from COVID.  Moreover, the spikes in the number of excess deaths correspond pretty closely to the times that COVID deaths have spiked.

https://www.cdc.gov/nchs/nvss/vsrr/covid...deaths.htm

The Economist does a nice job visualizing that (and other country's) data:
https://www.economist.com/graphic-detail...hs-tracker

They show the United States with a bit over 1 million deaths more than what one would have expected historically, which compares with about 800,000 deaths reported to have been due to COVID.  When you allow for the fact that early in the pandemic some people went undiagnosed (old people found dead in their apartments in New York) who might have been coded as a heart attack, plus secondary deaths due to the pandemic (for example, heart attacks that couldn't be treated adequately because a hospital was full, cancer that was detected too late due to loss of screening, suicides due to isolation from the pandemic and lockdowns, increase in drug overdoses, etc.), 1 million looks to be about right.

BC

[BostonCard]

Last point:  I have to take this moment to re-raise the issue of overreporting of Covid deaths with BC, who was adamant that all prior reporting was accurate (and frankly, did not understand why I was asking if the deaths were "because of" Covid or died "with Covid").

That mischaracterizes what I said.  Of course not every single case could be perfectly adjudicated, but if a patient dies from a heart attack, the death certificate will list heart attack as the cause of death, even if the patient happened to have COVID at the time, unless there is reason to believe that the COVID caused the heart attack.  It is impossible to adjudicate every case perfectly, and there is going to be some variability and biases among physicians, but I think the it will all more or less balance out and broadly you can trust the numbers of deaths that have been attributed to COVID.

An easy way to reassure yourself that the numbers are broadly accurate is to look at "excess deaths".  If you count up all the deaths, regardless of cause, since the pandemic began, and compare it to how many deaths you might have expected based on what has been reported historical (plus trends) it turns out that the difference corresponds pretty closely to the number of deaths that have been reported from COVID.  Moreover, the spikes in the number of excess deaths correspond pretty closely to the times that COVID deaths have spiked.

https://www.cdc.gov/nchs/nvss/vsrr/covid...deaths.htm

The Economist does a nice job visualizing that (and other country's) data:
https://www.economist.com/graphic-detail...hs-tracker

They show the United States with a bit over 1 million deaths more than what one would have expected historically, which compares with about 800,000 deaths reported to have been due to COVID.  When you allow for the fact that early in the pandemic some people went undiagnosed (old people found dead in their apartments in New York) who might have been coded as a heart attack, plus secondary deaths due to the pandemic (for example, heart attacks that couldn't be treated adequately because a hospital was full, cancer that was detected too late due to loss of screening, suicides due to isolation from the pandemic and lockdowns, increase in drug overdoses, etc.), 1 million looks to be about right.

BC