Originally Posted by SentientMeat
Imagine a large cube of ground driving around like a bus. A 10 ft fall is equivalent to being hit by that piece of ground at (hang on, v2=u2 + 2as, u=0, a=10m/s/s, s= 3m, so v = sqrt(60) = 8 m/s, and 10m/s = 22 pmh) about 18 mph. Now, many people might well survive the equivalent of being hit by a bus at 18 mph, but many won't. And as the bus increases its speed, it will kill more and more of those it hits (from the side, you'll note - supposedly the safest way to land). A 30ft fall puts the bus at around 36 mph, which is almost invariably fatal.
So from fall alone onto a hard surface, I could well believe >50% lethality from a height of less than 20ft. (48 ft = 40 mph!). The variability would come from the surface one fell onto: a wooden bus would be less lethal, and a grass bus less lethal still.
Hate to nitpick, SentientMeat, as sentient meat is one of my favorites, but if you are estimating death rates of pedestrians from mph, there is more information needed. You see, the sudden deceleration from hitting the ground is not exactly like the way people die when hit by a bus.
When hit by a car, the vast majority of people are not run over; they are run under. The lower legs break, sending them into the air. They usually strike the hood of the car, often with the back of the head impacting the windshield, "starring" the windshield, possibly leaving a few hairs in the glass. They then go over the top of the car. They are still alive, although with broken legs, and maybe with head pain from the nonfatal windshield impact. They die when they hit the ground. They die from head injury.
A person hit by a bus would actually be run over. There would be two separate sets of injury. The first would be when the kinetic energy of the bus struck the pedestrian. Mass times velocity squared. This is the closest approximation to the ground that your example offers. The likelihood is that they would incur lots of fractures, but not be killed by simple impact. If they were killed, the injury I have seen most often, in my autopsies as a medical examiner, is rib fractures, with the free ends of the ribs bending inward at the moment of impact, to perforate heart and lungs. This is a vanishingly rare injury in fatal falls.
If they are then run over, they incur compression injuries beneath the tire marks, which makes them, in layman's terms, squuushed. They are not squshed the same way that people who land after high falls are squished: very different sets of internal injuries. This is not what you wanted from your example.
When people are hit by the side mirror of a car, and spun off sideways, which is a better model of bus impact, they generally do not die even when the vehicle is going very fast. I see broken arms very often. Those who die, do so because they fall on their heads with enough mass times velocity squared to squush it.
We need a better example than a bus.