you are getting close to how science and testing work. Dogs need to be tested under the conditions taht they work.
there was a test carried out with expolosive dogs...its been posted here....where explosive residue was planted in a house and several dogs used to find the location. several dogs did...problem was there was never any residue palnted...the handlers cued the dogs
So in my example you'd also like it to be conducted double blind?
How would we be sure we could accurately replicate the exact conditions present in 5a at the time of the search?
How far do we go? And are you then suggesting that, because the use of cadaver dogs have not been scientifically tested these exacting, and probably impossible, standards, that their results should be inadmissible?
I'm the 'reasonable doubt' king - it's on my business card despite me having nothing to do with criminal defence, but even I'd be hard pushed to try that one, arguing on such tenuous a premise.