Perceptual filling-in only needs this kind of feedback if retinotopic visual cortex is a homunculus that instantiates subjective experience (the "isomorphic" theory of filling-in). An alternative is that filling-in involves more symbolic feature/object representations at a higher level.

PS: all raw data, processed data, and code are publicly available: doi.org/10.17605/OSF...

These results provide insight into microsaccades’ general mechanism of action in the visual system, and advance our understanding of the quite compelling illusion of filling-in. (3/3)

doi.org/10.1167/jov....

We found that microsaccades are indeed more effective when boundaries are closer to the fovea. However, they introduce equal delays across different levels of isoluminant color contrast, meaning that microsaccadic delays do *not* explain why less similar colors take longer to fill-in. (2/3)

that’s fair 😛 and maybe case in point that it’s distracting to neutral parties

totally agree except I think a focus on the pseudoscience claim is reasonable. attaching their arguments to such a sensational (and potentially unrelated) headline makes it a lot more difficult to take the points seriously

Of course, but maybe odd to tell the NYT editorial board that IIT work is not science.

I support the SOS. But 'pseudoscience' isn’t a call for nuance, it means IIT (& not competing theories) should disappear from scientific practice. If that’s what you mean, I’d argue you do have a responsibility to define pseudoscience beyond a speech-act. If not, then it’s not an appropriate label.