I want to be clear that the research CSC did is what's known as a secondary data analysis. They were doing a meta-analysis of previously published research available from other institutions and academic work. That's fine. It's a perfectly acceptable way of gathering whatever evidence is available.
The conclusion they come to, which is that this is an effective means of detecting drugs, is an interesting one, because essentially they are failing to make the distinction between a false negative and a false positive. This machine does not come up with false negatives. If the drug is there, yes, it will find it, absolutely. In terms of those conclusions, yes, it detects drugs.
The problem is not false negatives. The problem is false positives, where it's saying that, yes, there are drugs here when perhaps there aren't or the amounts are so small that they are negligible.
That's the issue with that sort of conclusion. It's just misleading.