The claim is that this evidence is weaker than the evidence against Guede by a factor of millions, and therefore deserves essentially zero brain time.

No, the evidence is weaker than it would need to be to overcome the prior improbability of "a rape known to have been committed by a random burglar in a flat was a product of conspiracy with a woman who lived there", plus the absence of Knox DNA is very strong evidence against her presence. The degree of the evidence against Guede has nothing to do with the bar - the bar is set by the prior improbability, not by the strength of evidence that Guede happened to leave behind himself.