Because scientists are woefully undereducated on the finer points of computer science. Does it (eventually) produce publishable papers? is all they care about.