Excess mortality is, sadly, a poor way to measure the impact of the virus.
Without the lockdowns, it could work. Using thrle virus to scare people (or legally prevent them) from getting necessary medical care means that we may have killed more people with the lockdown than the virus.
I agree, it's not great, but it sidesteps the whole "All deaths are reported as COVID, Therefore no numbers are to be believed!" argument so beloved by some. You'll note I carefully said
something killed a lot of extra people. I did not claim COVID-19 killed 300,000 people. Certainly some of them were suicides and/or murder brought on by stress, some were other health issues that were deferred and caught up to people, as you say. But a lot of extra people died this year, and Occam's Razor would imply at least a large portion of those people died of complications from COVID-19.
On your specific hypothesis, if people didn't go to the Dr. out of fear, and died of a different medical condition, we'd expect to see an up tic in non COVID disease deaths. I can find a Washington Post article from mid summer that I can not read due to not subscribing that claims to see an increase in heart disease deaths over the summer. (heart disease and stroke being the biggest single cause of death in the US, I figured any data blip would be seen there first), but I can't look at their data due to the paywall. I found an article from England and Wales showing an increase in at home heart attack deaths during the lockdown, but it's accompanied by a decrease in hospital deaths, suggesting that while some died that wouldn't have, some more just transferred their death to home from the ER. (
https://www.eurekalert.org/multimedia/pub/244268.php)
Total cause of death statistics tend to lag a couple years, so as you said, it may be a while before we really know.
But a bunch of extra people in the US died this year, one way or another.