Well, for example, I got a twinge of nerves on reading the following "red flag":

> Only focuses on a single group (e.g. women) to the exclusion of other forms of diversity

That is, if you make the mistake of going in too much depth on one particular aspect, then this is a red flag. I could easily imagine myself doing this simply because I have a tendency to answer interview questions depth-first rather than breadth-first.

I was also a bit terrified of "when I ask this question of an underrepresented candidate…". Some forms of group identity are outwardly visible; some are not. The author notes how sensitive this subject is, and then also predicates a fifth of the "discussion" section on the assumption that they can reliably make this judgement.

And on this "red flag":

> Indicates distrust of workplace inclusivity measures like codes of conduct, anti-harassment training, etc.

There was a good while when I distrusted all corporate training, simply because the only forms of active corporate training I had been exposed to were these box-checking legal cover-your-arse arrangements. If you've only experienced that kind of training, then of course you have a generalised distrust of it.

If you give me a list of criteria on which you are judging me, I will certainly be able to find ways in which I violate those criteria, no matter what the criteria are. (It's exactly the same problem as odd-one-out questions: there's always a way to make something the odd one out, so there's no right answer.) Hence the landmines: you can't know in advance which of those violations are actually meaningful to the interviewer.