A list of findings is more useful when the person fixing them can understand the situation. A score alone leaves quite a gap.
What would you look for in a testing workflow? How do you combine automated findings with checks that need a person to try the interface?