All models are wrong, but 1) some models are useful, and 2) you can't not have a model. If you don't have an explicit model, you have an implicit model.

I presented you a model; yes it's oversimplified, but it results in at least some sort of concrete answer; and you can actually examine where it came from and reason about how valid it is. You can make arguments for some other death rate or herd immunity rate. (I took 80% from the most recent article I heard on the subject, but previously I heard 60%.)

You also actually have a model in your head for how many people are going to die (or not die). But I don't know where it came from. I can't examine it or reason about it. If you were an old-timer who'd seen hundreds of pandemics, or a historical epidemiologist who had studied hundreds of epidemics, maybe your "gut feel" would carry some weight.

But since (AFAICT) you're a layperson like me, I'm not inclined to think your implicit model is going to be very accurate.