It devolved, because it went from what looked like a structured question with a finite expcted result, "Name the most common data structures" , to something more loose, a list of structures, any structures.

Further more, we agree, the candidate was trash, and I don't think the wording would have helped. But I do think a more precise question, or maybe a less loaded, or bias inducing question of "name some data structures". Namely, because "common" is subjective as I pointed out. If you are writing lisp all day, well, list are your most common. If you happen to be in assembly then registers are. So to be a better interviewer you don't want to taint the question with your notions.

It may not be unreasonable, but the question didn't ask that. Granted, it is unlikely the interviewer was skilled in interviewing and the ill-conceived question was born out of their own lack of ability in the the job they were thrown into doing. The reality is that very little effort goes into researching the science of hiring in the first place, and it is even less likely to find people who are both experts in technical matters and experts in what little we do know about interviewing. For how critical businesses claim the hiring process is, it is amazing how they will so often happily throw the first bumbling idiot they can find willing to accept the interviewer role into it like it doesn't matter.