No... doing your own analysis to verify the study is incredibly scientific. And the authors have provided you their code.

So find a bug in the implementation which can invalidate one of their claims. Lack of tests, is not a bug.