As metaphors go it's actually pretty good: people often have very specific, very individual, and entirely valid reasons to not eat something. Just as they can have the same for experiences under different conditions. Change the conditions, and you might get entirely different outcomes.
What I'm saying is, the one-shot impressions from interviews are horribly inaccurate. There is simply not enough data, plus the measuring instrument (interviewer) is flawed also in nonreproducible ways.
A simple thought experiment: let's say I build a machine that measures confidence and ability to read the room. Would it select different candidates than you?
True, I meant it as a metaphor for the kinds of new experiences/ideas one encounters day to day in the workplace.