I've no idea who surveyed the traffic, but perhaps they asked the vehicles the wrong questions?
Only if the vehicle was registered at the wrong address. They used NPR backed up by other surveys and interviews which showed a significant percentage on the trunk road was in fact local traffic.
