The point is that they are parameters of relevance to observer.
I work in medicine sometimes work with box-plots for this reason. The questions "what is the 25th percentile outcome" is perfectly legitimate