I feel like your statement reduces to "people are not affected by evidence."
There is a category of people for whom that is true, but is there not also a category of people whose beliefs and actions are affected by evidence?
That may be the case now, or maybe it only was in the past? What will we do when ALL true evidence is indistinguishable from fake evidence?
(By the way, I think you mean "provenance," and as I've edited my post and bio blurb to explain, I too am against methods that disclose the identity of the witness. Rather, my methods would help multiple witnesses to an atrocity [such as police firing live ammo at nonviolent protestors] to corroborate one another's accounts before identifying themselves. They could possibly even effectively discredit an official, faked account without identifying themselves.)
I mean, at least in the US, trial lawyers bill by the hour, which, as I see it, is already a conflict of interest on a meta level. This system is pretty inefficient. What if we could prevent most "he said, she said" situations (and, probably, billable hours) by making recording of all one's experiences the norm, while also safeguarding the privacy of those recordings until such time as a dispute (such as a court case) arises? (This latter, especially in the context of doctor-patient relationships, was the original root, circa 2004, of my line of thought. The idea then was "universal surveillance by private individuals of their own lives, encrypted by default, with proof of existence, optional decryption, and sharing.")