About 300 non-occupant (i.e., in the street or on the walk) children a year die from car crashes each year in the US. About a third of those are in school zones. This means two things: one, a solution to force slower driving in school zones would not prevent the majority of fatalities, so if that solution was counterproductive outside of school zones, it would reasonably increase fatalities. Two, 300, or 100 in a school zone, suggests the vast majority of car-child interactions are not fatal or even injurious. A solution that prevented a couple of the existing fatalities could easily turn more of the much larger number of safe encounters into dangerous encounters. And this is plausible because right now, people are watchful and quick to slam on the brakes when they see a child, even out of the corner of their eye. Teaching them that an area where children walk has a “child” they can ignore would weaken that reflex where it should actually be strengthened.
I understand you are concerned about children as well, so I would advise looking into policies regarding parking on the side of the road. Roundabouts in neighborhood intersections near schools also have demonstrated a natural slowing effect that does not seem to have negative externalities beyond the cost to implement the roundabout.
