You cannot directly compare two populations without accounting for differences in size. Doing otherwise is very bad data science. That's why it's per 100,000 and thus takes size out of the equation. Which is good. That's a confounding factor that is trivial to deal with. Given my previous observation that many US states have populations directly comparable to other countries, the "comparing states with countries" complaint goes from vaguely plausible to inane immediately.

Thank you for defining how averages work?

Source: data engineer