How about per fly versus per drive? This weeds out two issues:

- Most people don't fly often enough to justify Statistics significance (I for one only flied maybe less than 10 flights in my whole life)

- One flight is going to cover a huge amount of mileage anyway

Edit: Just realized that issue 1 is not an issue, we are going to do an average here anyway, so not individual.

Living rooms are also hilariously safe. And we spend a lot of time in them.

Is there actual reason to think they are less safe per hour of time being spend in them as OP claimed?

Stairs?

Spouses

Per passenger mile is arguably not the best denominator. People choose planes because they are going long distances. Consider whether a better denominator would be per passenger trips. A 10,000 mi trip halfway across the world could have the same weight as a 2 mi trip to the grocery store. Or per hour travelled.

By these metrics commercial flying isn't as safe as you think.

If I’m going somewhere in the continental US, my choices are to fly or drive. I’ll be traveling the same number of miles either way, so the relevant comparison is indeed per mile.

That doesn't explain anything. Specifically it doesn't explain why in our comparisons the destination is fixed. You could decide for this weekend trip we're going to budget a max of 3 hours on transportation. Should we take a three hour flight to a different state or should we drive for three hours to a closer destination?

This just isn’t how normal people behave. They are starting with a destination (relatives, national park, etc.) and then choosing how to get there.

That's not how normal people behave. If people choose a faraway destination first they are implicitly assuming they will fly there.

except 10,000 miles by anything but air isn't a single trip, it'd be multiple trips and involve a boat so that's not really a fair comparison either.

Furthest you can go in a straight land on land is about 7000 miles :).