I've been fairly disappointed with the analysis of first doses first. If someone were to recommend half-courses of antibiotics, I feel like a lot more people would instantly recognize what could potentially go wrong here (and for which we will have 'no evidence for' until it is too late). The assumption of first doses first is that partial protection is better than all or none - but with evolving organisms, this is a pretty tendentious assumption. It is not at all implausible that this strategy will select for vaccine-immune strains.