first of all you said there was nobody immune to prejudice, but when called on proving it you turn it into a question about my own immunity, which I never claimed to possess.
second of all this seems like a variation of that old interview question - tell me a reason why we should hire you over any other candidate that had your same skills but had X (I was asked once why should we hire you instead of someone who speaks our language better) to which I replied obviously if you have someone who has all the same skills but one thing better than me you should hire that person.
Going back to the problems that were initially under discussion; The problem that is supposed to be addressed is not candidates who have all the same skills and abilities and how to choose one because at that point you might as well just draw lots. The problem is candidates who have superior skills but are not taken, candidates with inferior skills but right social markers being preferred.
> it'll be a contorted mess of statue plus case law plus HR policies plus things you can clearly articulate in advance in the job criteria
HR policies and clear articulation of job criteria are things the company controls, but in your scenario they seem to be some insurmountable barrier put in place to catch them out. An organization that is damaged by its own HR policies and job criteria should do something to fix that.
>and prove to some jury against a plantiff with a sob story.
If you chose a candidate from a list of candidates that all were exactly the same qualified it seems doubtful that the sob story will be much good.
Finally:
candidate number 1 doesn't seem likely to me because my prejudices tells me the candidate with face tattoos AND ear gauges is most likely a member of Earth First, and is not going to roll coal but will probably have gotten to the interview on skateboard or something else generally environment friendly.
Candidate number 2 so you're saying HR made a mistake and he did not pass the speaks English requirement adequately.
I don't think I would care much about differences between 3 and 4.