I find myself thinking of the differences between a single-shot "Prisoner's Dilemma" and the "Iterated Prisoner's Dilemma"; that is, conclusions about any particular choice may not necessarily apply when dealing with a large number of similar choices. That is, even if, in the listed examples, the QALYs indicate a particular choice is the best one, that does not necessarily imply large-scale social policies should result from that conclusion. Or, put another way, the QALYs of a nation having a strict anti-abortion policy or a more permissive one don't necessarily seem to be closely correlated with the QALYs of an individual facing a choice on the topic.