Comment by Turing_Machine

This seems wrong. If they're using code to arrive at their findings, it should be high-quality, no less so than their lab technique. One can lead to bogus results just as easily as the other.