It's also a factor that acceleration/deceleration in an ICE kills mileage. Highway tests maintain a constant speed. If you ran the same test at 35 mph, they would get much better mileage than at 55 (or 70)

Idle losses are real but not very substantial in a modern engine compared to the bigger factor you're missing which is that in city driving tests there is a lot of speeding up and slowing down, ICE vehicles throw away all the energy used to slow down as heat in the brakes which makes city cycles particularly inefficient while an EV captures that energy through regenerative braking, dramatically reducing the net cost of those momentum changes.

Also regenerative breaking is not useful when you're maintaining constant speed on the highway, but a huge leg-up in the city

That makes sense, thanks.

So does the EV9, but it exceeded the EPA range in the same test.

Aha yes, but what about other EVs?

That doesn't explain why ICE vehicles are much better efficiency-wise than EV ones at higher speeds, just why EVs don't do so well. Another post responding to me addresses it well.