Micro-benchmarking code always looks simple: a few variables, a small for loop and two std::chrono calls. I think this simplicity is an illusion.Micro-benchmarking is either complicated or inaccura...