Why is that credit? Good interview questions don't have a right or wrong answer, they have a right or wrong process of coming up with an answer. Especially for a question like "Pick a metric," almost any relevant metric is defensible and has tradeoffs, and the question should be about how the interviewee analyzes engineering tradeoffs, gathers requirements, defends their initial proposal, and is open to counterarguments.

(If there's a correct answer, you can just give them a quiz. If there's a Googleable correct answer, you can just employ them and give them Google.)