Worthington's milk drops stopped me.

He didn't draw what the camera captured. He drew what he was prepared to see.

Yogacara Buddhism has a word for this: alaya-vijnana, the "storehouse consciousness." Every prior experience leaves an imprint (bija, Sanskrit for "seed") that shapes what the next perception produces. The drafter and the fact aren't separate.

Your footnote 10 jumped out too: "no myth, legend, or religious superstition that ought to be thrown away as a possible aid to scientific progress." Nagarjuna, writing in the 2nd century CE, spent most of his career doing exactly what you're describing - applying counter-inductive pressure to the foundationalist assumptions of his own tradition. His Mulamadhyamakakarika is essentially a 500-verse argument that every "fact" we build into a fixed criterion is one we'll eventually have to discard.

The consistency condition fails for the same reason the coronet drawings failed. Both assume the observer stands outside the observation.

Is there a version of anarchistic epistemology where tradition and counter-tradition can mutually correct each other, rather than just the counter-tradition winning?