I don't see why that is a bad question. A reasonable answer is selecting from a table. Of course it depends on many factors. I often ask questions like this just to get the candidate to tell me why there is not an absolute answer.
Yeah that's the thing. I was talking about computed columns in tables, materialized (or indexed) views, about measuring stuff by experiment and about measuring stuff that actually matters but the interviewer seemed like he wanted to hear a simple answer
"Which is faster: select from a table or from a view?"