[philsquared_]

I mean you probably could run a study of caffeinated vs un-caffeinated drivers. Or Modafinil vs Un-Modafinil drivers and I imagine the results would be: the ones with the drug perform better than the ones without.

Expanding on this: I am sure more fatal accidents occur late at night than during the day. Do we bar night driving because of the risk?

Bigger cars cause more fatal accidents than smaller cars. Bar bigger cars from the road?

[ceceron]

You can imagine the drug would help, sure. But that's it – your imagination. But yeah, if we assumed the Modafinil was indeed a proven "cure" against sleeping while driving, and it was given free, then the situation would have a similar structure as the one analyzed by the author. Obviously, they still differ significantly: a) you would have probably to take Modafinil before each drive – vaccines have a more permanent effect b) it is easier to track vaccinations than taking a drug, it's much more manageable.

This is not only a risk management game – your simplified "absurd" examples are of the kind: "author wants to bar risky situations, why not bar big cars, they're risky". One has also to weight various social factors/costs. If barring bigger cars from the road would be beneficial for society, then yeah, let's do it, but I wouldn't bet on that. The vaccinations have some added costs (production, distribution) and risks (side-effects), but the question is whether the cost/risk of the longer pandemic isn't even bigger (AFAIR it is). I don't say, that OP is right that people should be forced to be vaccinated, but I don't find it absurd given that we've already sacrificed many other freedoms for the society. It's like @mayewsky writes, we already bar some drivers (because of age/health state/previous offenses against the law) from driving. It's nothing new and it's not controversial AFAIK.