[DenisM]

Run a benchmark to compare performance of my product before and after making a change in code.

[gus_massa]

I run it 5 times and drop the two extremes. If the middle tree don't overlap, I'm probably fine. Not super precise, but good enough and easy.

Also, add a loop so the benchmark is like 1 second long, I'm old enough to be worried about anything smaller than 1/18 seconds.