This "average life expectancy in the ancient world" thing is nothing more than an illustration of when to pick the median over the mean because the data is skewed by outliers.