If it’s like (or based on) the ICCT report I read the other day when this started blowing up there’s not really enough data to draw a conclusion. Anonymous test performers on uncontrolled routes with cars of differing mileage/maintenance.