These devices probably cause < .1% of fatal pedestrian accidents
Percentage is meaningless without context. The stat you’re actually looking for is pedestrian deaths per mile. And it’s probably quite bad for these vehicles because they explicitly commingle with pedestrians.
Cars don’t spend very much time on parts of roads that have pedestrians on them, and when they do, there’s signage or traffic lights to help. Cars also have lights to help drivers see pedestrians and help pedestrians see cars, and generally make a lot of noise. You get none of these benefits with personal motorized vehicles. (Well ok, a scooter probably comes with some lights, but they’re probably also small and shitty and unregulated, so they don’t really count…)
It’s a cat and mouse game, except the mouse has effectively infinite lives.