Well, would you agree that the phrase "shut that voice out, ruthlessly" could plausibly be interpreted as meaning "ignore that intuition entirely"?

Tabooing "ignore", I do not think "that intuition constitutes literally zero Bayesian evidence of guilt" is a reasonable interpretation of my phrase. "Your analysis of this case should be for all practical purposes the same as it would be if you didn't have that intuition" is more accurate (where "practical purposes" include things like deciding whether to regard Knox and Sollecito as "innocent" or "guilty", and how strong a level of rhetoric to permit oneself in expressing one's conclusion).

I think it is clear that this post is not about how to distinguish between the kind of extremely small differences in confidence levels that would be entailed by correctly factoring that intuition into one's calculation. Remember that the entire context here is a catastrophic failure of epistemic rationality -- with horrible consequences -- that resulted from people paying too much attention to these types of intuitions. They would have done much better to literally ignore it entirely. (If it actually raises the probability of guilt from 0.001 to 0.002, it would have been better for the police or jurors to have ignored it entirely and estimated 0.001 than to have taken it into account and estimated 0.99, as they effectively did, or 0.5+, as many others have done.)