Well, you're arguing (and I agree) that poor code quality can be identified, but not necessarily that it can be measured. The symptoms you list are partly caused by poor code quality, but not entirely either.