To a degree, it doesn’t matter. According to
Provisional weekly death counts: Interactive tool we haven’t been experiencing catastrophic levels of deaths overall during the pandemic. It looks worse than it really is on this graph because the graph is truncated at 4500 deaths and above. Truncated graphs are not at all unusual or invalid. They do, however, emphasize change rather than showing absolute values. This can be useful to show things like the degree to which deaths in Canada tend to happen more in the winter than in the summer. It also clearly identifies the spike when Covid first appeared. Note that the chart shows ALL deaths rather than Covid deaths. The total would include influenza deaths which would be near zero for 2020 probably in large part due to Covid countermeasures. The total would include suicides which are apparently significantly down in 2020. It would also include opioid overdoses which are apparently way up. These could well be counted as Covid countermeasure deaths. The reference article
Staggering Surge Of NYers Dying In Their Homes Suggests City Is Undercounting Coronavirus Fatalities contains a quote saying ““[People] may be dying because of reduced care for other non-COVID diseases” like diabetes, heart attacks or other chronic conditions, Redlener said. “Those to me, should be somehow tallied as we’re looking at the death toll of COVID.” To my mind, these deaths due to reduced care should more reasonably be counted as a death toll of Covid countermeasures rather than Covid.
It is also interesting to note that deaths during the Covid spring peak were not hugely higher than the worst two weeks in 2018.
For comparison, I have created an untruncated graph for 2018, 2019, and 2020 deaths that looks somewhat less alarming. We must note that every point between 2018 and 2019 on the one hand and 2020 on the other is an actual person who died whether unrelated to Covid, due to Covid, or due to Covid countermeasures. The untruncated chart looks much less alarming, though.
View attachment 64745