I know it is a rather old topic, I just stumbled into it and was wondering if I understood correctly the problem.
From the documentation:
A piece of code is considered duplicated as soon as there are at least 100 duplicated tokens in a row spread across at least 10 lines of code
From a high-level perspective, the problem is that you have to call the same big sp_addextendedproperty stored procedures in multiple files/places, and there is no other way to avoid doing it.
In other words, duplication detection is correct, it is just that you cannot do without it. Did I understand correctly?