It's a hard call to make either way, because you have to make an assumption about how many of the currently uninfected people will eventually get infected with

It's a hard call to make either way, because you have to make an assumption about how many of the currently uninfected people will eventually get infected with, say, Delta (or any other variant) in order to calculate the tradeoff. If a more infectious variant just gets to people faster, but Delta (or whatever other variant) was going to get to them eventually, then that changes the tradeoff. A better argument against it would be that there is a huge unknown, and the very uncertainty about something which cannot be undone, is a reason not to do it.