Better Descriptive Statistics for Low-Latency Profiling

How often have we been tasked to characterize a given event or container - most items by ourselves because who does not like to run a benchmark - but then get results that do not tell us anything?

Ma…