Even for societal aggregate worries it’s only worth using days as the denominator if the potential interventions you’re considering scale by days and not person-days. For example “stay-home on Halloween” or “move Halloween earlier” has costs proportional to the latter.
Nicely done!
The question is what should be the denominator of the risk?
For societal aggregate worries: Day? Sure.
For personal risk: Day? Nope. Person-day? Yes!
Even for societal aggregate worries it’s only worth using days as the denominator if the potential interventions you’re considering scale by days and not person-days. For example “stay-home on Halloween” or “move Halloween earlier” has costs proportional to the latter.