Evidently there's at least one group of developers who mainly derive their decisions from irrelevant and often incorrectly conducted micro-benchmarks.