If this was at all realistic, we'd also see similar effects in people in other fields - bus drivers more likely to crash before lunch, etc. That we don't see su

If this was at all realistic, we'd also see similar effects in people in other fields - bus drivers more likely to crash before lunch, etc. That we don't see such an outrageous difference implies that the original study is not tracking the relevant variable when it looks at before/after lunch.