This is the best argument in the series against desert by far and I think sort of settles the debate unless I'm wrong about something important
To me, it honestly doesn't seem counter-intuitive at all. This just seems largely a result of the framing and perhaps heuristics bias.
Nice post! A serious challenge to the seriousness of desert. A couple of thoughts:
> "if the most virtuous possible person and the most vicious possible person have already suffered an unfathomable amount, then the marginal badness of adding a bit of extra suffering tend towards being equal"
This seems defensible, by analogy to a criminal having "served their time" and now their slate is clean. Unfathomable suffering is already far more than anyone could really deserve, and so any extra suffering is strictly undeserved.
> "If there were lots of vicious happy people, that doesn’t seem like a terrible world, but it would be on this account."
fwiw, that seems like a bad world to me! Not sure if I'd endorse the verdict on reflection, but it at least isn't *obviously* wrong.
In general, I think the "better to be worse" result is avoided by thinking that virtue/vice is more axiologically significant than desert. But this leaves open how big of a deal virtue/vice is. It could be that (i) raw welfare almost always takes priority over (ii) greater virtue, which almost always takes priority over (iii) desert.
To sufficiently minimize the impact of desert, as you say, it probably needs to be capped. Which suggests that, to a first approximation, we can just ignore desert in practical ethics. Whether it's non-existent, or merely negligible, doesn't hugely matter (aside from sheer philosophical interest).
Desert matters sometimes and cases like this just aren’t those times