As others have said, and I would agree, that title isn't misleading as any day is "user data" and your argument revolves around the semantics of what "personal data" is (or isn't).
Beyond that I think it's fair for folks using Sentry to be disappointed in the decision. It's very hard to classify data types to be excluded and much easier to go down the "inclusion" route. But in this case Sentry can later argue that, whoops, these data were incorrectly classified because the user interpreted our Rube Goldberg machine incorrectly. Par for the course with respect to anti-patterns.