Running the benchmark commands with the name of a non-existing test will happily produce no results instead of yielding an error. While this is somewhat quick to find, the better experience would be to yield an error to the user that the test does not exist.
Running the benchmark commands with the name of a non-existing test will happily produce no results instead of yielding an error. While this is somewhat quick to find, the better experience would be to yield an error to the user that the test does not exist.