The article (page 41 to 45, in July/Aug 2021 issue) has a much broader scope than what I will cover.
Here is the text of a table, "Table 1: Initial Ad Hominem Rubric" :
(The format will be changed, but the text will be the same, except "Type" and "Brief Description", will be left out.
Attack on data/claim because the source lacks expertise or has been formally disavowed by some institution
Attack on data/claim because the source is emotional or blinded by passion
Attack on data/claim because the source is unethical or of bad character
Intelligence/ bad thinking processes
Attack on data/claim because of the way the source thinks or doesn't think ( e.g. stupid, close-minded, ignorant)
Attack on data/claim because the source has a vested financial interest
Appeal to negative non-financial motivations
Attack on data/claim because motives of the source are bad for non-financial reasons
Association with bad people/groups
Attack on data/claim because the source is a member of an organization that is known to be bad or because the source associates with people known to be bad
Attack on data/claim because the source is bad in a vague/generic sense(e.g., a jerk)
Attack on data/claim because the source engaged in a bad action
Other negative qualities
An attack on data/claim because of something negative about the source, but doesn't fit into any other categories above
End of table text
This personal attack issue is something that I have a mixed opinion about. I feel that ad hominems should perhaps be labeled as such, as an editorial note, in the body of a person's text.
The article has a broad scope, but to keep with my limited scope, I will reference the article's example of Andrew Wakefield's study which showed a significant association between children taking the MMR vaccine and autism spectrum symptoms in the subjects. The Skeptical Inquirer authors pointed out that an example of ad hominem attacks on his research would involve the comment that Dr Wakefield's medical license was revoked. The "argument directly targeting Wakefield's claims might focus on the small sample size, inaccurate data reporting, and lack of replication of Wakefield's now-retracted Lancet article".
I think ad hominems are extremely harmful to information debates, when they take up more than a very very tiny sliver of a post. I feel that the moderators should have to comb posts for ad hominem comments, and highlight the identified text in a color code. It would help keep the discussion targeted toward proper argumentation.