That could be true, but it isn't, since we don't always know at the individual level if a specific tumor is harmless at the moment of diagnosis.

You could say the same thing about diagnosing cancer via astrology.

A test that with low specificity is simply not a useful test.

> we don't always know at the individual level if a specific tumor is harmless at the moment of diagnosis.

You didn't know prior to the test either.

We don't know if a tumor is harmless, but we do know that treatment has significant harm. Radiation and chemotherapy are nasty. We also know that the stress of a positive diagnosis has significant harm.

First do no harm.