I have a scent discriminating search and rescue dog who can find specific scent in complex environments. I've challenged her at the entrance to a Walmart to find the scent of one specific individual through the morass of old and new scent from the thousands who have walked through the entrance and found her capable. I'll bet that she would not be surprised by your results. I'd be very curious if the live microbes on your filters can be used to scent a dog to track a specific individual. Up the ante in the next experiment and send out a negative control decoy human in addition to the true subject to verify specificity.