A note on a quantile-based probabilistic mean value theorem