This is the key of this two-pronged approach, one commenter can bury the data driven comment in source rejection (without being beheld to prove a counter point, since the asserter has the burden of proof) while the sister comment can drive the more approved comment unchallenged. Of course we really know, in many cases, the two separate commenters are advancing the same line of opinion, but using this split strategy both are compartmentalized in their burdens.
Although, the truth is, the scrutinizer is rarely offering counter sources of their own, which they of course are under no obligation to provide. But barring that, we're left at worst with "I don't know" which is a terrible standard under which to assume the word of the wife is predictive of guilt, thus even if all the sources are rejected you leave from a practical perspective no off no better than you started in predictive guilt.