Didn't they do a trial of this in the late 60s and actually there was a reduction in road traffic accidents because evening rush hour was lighter?
Yes. Or no.
The trial of permanent BST showed an overall reduction in road traffic deaths (in an era when there were a lot more deaths on the roads).
However, the children who died in the darker mornings had names and faces, and each one was a potential "had it been lighter, would they have lived?" debating point.
While every child who didn't die in the lighter evenings was a statistical abstraction, without a name or a face.
Making these sort of risk tradeoffs is not easy.