Worst possible case regardless of likelihood is not a good metric. I'll bet the average drunk driver, and indeed the average sober driver, causes many more deaths than the average unvaxed unmasked person in a post-vaccination world.
In the 2 years of covid we have had 4.25 million deaths with still 90% us and 97% of the rest of the world left to catch it.
You only need a unmasked person to have an R value 1.64 times higher than a vaxed/masked for the proportional deaths to not reach parity even with 90% people vacinated.
That's completely ignoring the fact that driving is actually a productive activity with a goal. Where as most of the reasons people give for not wearing a mask are either easily debunked or work just as well as a retort to someone asking you "why are you continuously punching yourself?".
https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/burd...
And most antivaxers and the unmasked are welfare cases with nothing to do but look at reddit, and so are a burden on the state and a direct cost to me as a taxpayer even when there isn't a pandemic. Hey, maybe this will work out great in the end.
And how many of those were of fully vaccinated people?
Also the dynamics of how far any given infection would spread are very different in a post-vaccination world.
I'll admit something. I happen to love driving twice the speed limit after a few drinks. Feels great. And I indulge that sometimes on country roads when there's no one around. I know I'm a better driver after a few beers than most people are sober. If it were up to me, there would be a test, and a rating on my license that let me drive with a higher BAC than other people. But you know what? My personal pleasure isn't worth the damage of taking someone's life, or of entitling everyone else in society to do the same. So, on the whole, it's obviously better that it's illegal. Even though it infringes on my personal freedom and pursuit of happiness.