I agree there may be quite a large range but just to say that can still be useful.
I think its crucial to start denormalising all the costs and externalities of car focussed transport policy. Motornormativity means policy makers and general public internalise costs of progressive infrastructure and are blind to the huge costs of the status quo.
So even being able to pin a wide range on it can be helpful. Not for financial costs but for emissions I was able to show even for the lower end of a wide range of additional hard-to-quantify emissions for scenarios that didn’t drastically reduce private car usage as well as electricify would blow past thier carbon budgets.