I would agree with everything you said below IF I had agreed with your premise. I don't. First, 'owning' spectrum is a misnomer. You can't actually own spectrum. You can be given a license to transmit in a certain frequency, but you don't own. Anybody could interfere with your broadcasts if they wished, with only the fear of legal ramifications. Granted, you can broadcast in a certain frequency all the time continuously, thus hogging the spectrum (essentially UWB floods large portions of the spectrum, however it uses techniques such as OFDM to keep things working). You can't however broadcast in all spectrum all the time, unless you have a very fancy multi-in multi-out type transceiver. Besides, current battery technology wouldn't let you use this type of power. On top of this, the whole thing would be pointless because anybody can broadcast at that same frequency and at that same time resulting in the degradation of your signal. Spectrum is finite, simply because there is a range and power within which we can broadcast and not cause nasty side effects like cooking everything in the broadcast path. However, within this range, the spectrum is infinite (considering time). The point of cognitive radio technology is it actually negates anybody's ability to "squat" on or own a piece of spectrum. The smart devices will see this spectrum as not available, and find the next white hole, which invariably must exist. A more effective solution to the economics of spectrum is to have a system where by the smart devices themselves negotiate for spectrum, with automatic transactions. For example, if you want to send your powerpoint project in to work from your pda, the pda will negotiate with various providers (cell, Wi-fi, or whatever else exists), and then pick the cheapest and most efficient alternative based on what it has learned about your needs and habits.
My point here is the spectrum itself will become a commons. The real economy will be in helping you get the information off wireless and on to the wired backbone.