You have already identified completed, active and withdrawn properties as separate groups; what remains unclear is whether the two neighbourhood samples use consistent boundaries. I agree that 66 days is useful, but only as a feature of this selection rather than a rule for London.
I would...