It would help if instead of just the average the author plotted error bars with a measure of the deviation: standard deviation, coefficient of variation, etc.

It may very well happen that the dispersion of the travel times is high enough to render fine differences in mean statistically meaningless.