I think it's kinda interesting still, in that it shows people are (must?) drive so much. But yes, agree that per cap seems like the wrong statistic for any kind of safety.

Maybe more fair would be transit fatalities per mile traveled (any method)?