Frequency would be scarce in highly concentrated areas in an open-frequency region, but it would be less so than if you artificially restricted access. What would naturally follow is that the demand would distribute itself (within the realm of possibility, ultra low and high frequencies have either range or bandwidth disadvantages that make them ill suited for, say, digital packet signaling).
I also wonder what kind of innovation you might see in an open band environment in terms of wifi - if you weren't limited to the few bands of 802.11, maybe routers would pop up that dynamically enable multiple transmitters (like how a modern router would have multiple antennas to transmit in parallel, maybe you would see many-hundred low power antenna radios to interface with devices over the largest clear possible frequency range available?)
But for use cases that require consistent signaling, or really high throughput, enabling radios that utilize vast swaths of frequencies in parallel can increase the throughput.
I work in the wireless networking field, and my and my coworkers' experience is actually the complete opposite in high-density environments: there is an abundance of radios, and a scarcity of frequency.
Which, BTW, is one of the reasons wired networking consistently outperforms wireless - you have one and only one transmitter (in a full-duplex connection) on a given medium.