Thanks so much, Nir.
On your first point, as I say, it doesn't haven to be completely random: we can adjust someone's odds upwards if the intervention they need is especially cost-effective. The exact adjustment in odds for very cost-effective procedures can be decided by the policymaker. This allows a compromomise between pursuit of fairness and pursuit of maximum aggregate health.
On your second, I'm not convinced rightaway but suspect there's something here! Let me think about it.