It's worth pointing out that in the origin of the "10x developer" term, it's relative to the worst performing devs, not the average.

Also, you're not guaranteed to have an example 3 standard deviations above the mean. It strongly depends on your distribution and sample size.