Well, this study doesn't prove they're unreliable. Even if the false alarm rate was 1000%, they would still be quite reliable, as long as the rate of false negatives was low. Who cares if they have to search a few extra people out of a million people to find that 1 in a million with a bomb? That would be extremely efficient. Well, as regarding bombs, at least. Drugs might be a different tune.
The TSA has "red teams" that routinely smuggle bomb and gun items through security checkpoints. That's their job. I don't know if they test canine units but TSA security has a fantastic failure rate to catch these smuggled items.
How does a false positive rate of 25-50% equal "a few extra people out of a million people" and how is such a high error rate "extremely efficient"? What's the point of substance-sniffing dogs if that's your mentality? Just do a 100% scan on everyone, security theater be damned. Lets bypass the Constitution and go straight to fascism and chock up the inconvenience and intrusion as a casualty of security. That's the same mentality fueling the rise of paramilitary police raids. Who cares if a few toddlers and old women are killed by police serving no-knocks on wrong addresses, false informant information, or victimless crimes like smoking pot or VFD vets playing some poker? Hey, as long as we nab some bad guys somewhere in the mix, right?
There was a study done last year on false alerts. There were tests designed to trick the dogs and the handlers. The handler's were twice as likely to subconsciously trigger the dogs on these tests. That suggests the human element is a bigger issue than the canine element. Sounds like operator error, and it appears it isn't necessarily malicious in nature.
http://www.economist.com/blogs/babbage/2011/02/animal_behaviour
In Australia, they are getting an 80% false positive rate. 11,248 nothings out of 14,102 searches.
http://www.smh.com.au/environment/a...ng-four-out-of-five-times-20111211-1oprv.html
And here is a fairly recent article on an incident where the drug dog played a role:
http://www.huffingtonpost.com/2012/...ped-searched-illinois_n_1364087.html#comments
I don't think alerting to gun care products would be useful because many gun care products are simply re-branded commercial or industrial products. There isn't really a gun lube company out there that has it's own processing plants, men in white lab coats swirling beakers, and a room full of scientific and mechanical testing gear to produce enough volume sales of 4oz bottle of $10 gun lube to sustain such an outfit. Most of these companies buy a base petroleum product and add a few additives to make it their proprietary blend, slap on a label, and call it good. Check out the MSDS sheets some time.