If you want to know if the impact of COVID you could look at a metric like excess mortality which there's pretty good data on, then compare the excess to the total mortality. Usually it's calculated so that any excess deaths is pretty rare which is what makes it a good metric.

With year on year increase in deaths I'm not sure I could convince myself it was significant unless I had a chart and a calculator in front of me. A year over year death increase, even a large one, seems like it would occur pretty often. Even a large increase in the increase doesn't sound like it would be significant -- wouldn't that be pretty sensitive to the shape of the population relative to age? And I'm not sure what that looks like off the top of my head. There's a lot to adjust for, and you need to be sure you're not just rolling dice at the end with the numbers.

My point is it's not really a good metric. There's plenty of reasons why it would jump around, and I'm not smart enough to adjust for all the necessary variables in my head. It's not obvious how much we should expect it to jump around. But maybe others here just know their US demographic data better than me.