I’m afraid my original question probably has a complex answer.
First, the “badness” is a distributed variable. That is, some people quickly die, some people never notice symptoms, and many people are somewhere in between. Each disease produces a distribution of severity, not a single value. “Worse” would be much more simply answered if both distributions were similarly wide but they were relatively displaced, but I don’t know if that’s true. It’d be great to see two distribution curve graphs with “badness” on the horizontal axis. But, there’s not just one single measurable “badness” variable, so maybe a practically ideal answer would be percentages for “no symptoms”, “mild”, “severe enough to want to stay in bed”, “hospitalized”, “intubated”, and “death”. Something like that.
Second, apparently outcomes depend strongly on vaccination status, so perhaps we’d need to show different distributions for the different levels of vaccination, and maybe the different vaccines.
But this all supposes we have complete information. Many people have these diseases but never have tests confirming them, many cases aren’t reported, and on and on.
Flu is worse than colds. I guess I know this, but am not sure how. Maybe there’s not much better answer for my question.