We'll never have precise figures on all deaths from covid. E.g. an 80 year old has a bad cold which was actually mild covid, recovers, and then a couple months later is out pruning a tree and collapses and has cardiac arrest. It may be impossible to say if the earlier bout of covid was a contributing factor or not. Or if he had sinus tachycardia due to an overactive thyroid which hadn't been treated, or etc.
That is why the statistical concept of Excess Deaths is useful with this kind of pandemic. It's not an Ebola outbreak, where the deaths are unequivocal, easily distinguishable from other COD. But ED gives a broad metric on the overall effect on public health of a pandemic, by comparing an average death rate to a current time period.