Good read 👏 your benchmarks however provide no real value. As you state yourself the tests are to small and the results fall within error margin where one or the other could be faster just by accident. Why not create more sophisticated benchmarks to prove your points more objectively? I think that would improve your story greatly 🙂