Let's say I have two functions that do the same, on different ways:

final class MathUtil
{
    public static function heavyCalculationWayOne(): float
    {
       // ...
    }

    public static  function heavyCalculationWayTwo(): float
    {
       // ...
    }
}

And then I benchmark both with the same revolutions and/or iterations.

final class HeavyCalcBench
{
    function benchWayOne(): void
    {
        MathUtil::heavyCalculationWayOne();
    }
    
    function benchWayTwo(): void
    {
        MathUtil::heavyCalculationWayTwo();
    }
}

It would be nice if one could somehow mark one as the baseline for the others inside a single phpbench run process. To easily see on a single glance if or which one performs better. Currently I have to evualate the runtimes, memory, standard deviations, ... of both manually to each other. Basically what regression testing is doing but instead inside a single run on method level.

An example use-case is when a computationally heavy task is to be solved and different devs try their approaches - to then decide which makes it into the final product. Another one: on new PHP major versions, to check if the PHP interpreter can now handle both ways roughly the same as fast, so a real-world application can now use the better readible/maintainable method instead of the unreadible-but-more-optimized one.