If planning a decade ahead is hard for the regulator, what about the system designer. We are now expected to plan, research, test, manufacture, and sell a system with no certainty about the future availability of the bandwidth we will be expecting our customers to use?
I'd want to be guaranteed access to a frequency band for at least 30 years before bother.
I can recall the 1950s and 60s, when allocation was done first by considering the physical properties of the band (line of sight, ground wave propagation, multi-skip, multipath immunity, rain attenuation) and then deciding what class of service it would be suitable for. After that there were international agreements between regulatory authorities. Only then were the people who wanted to make money allowed into the discussion.
One of the things that has rather horrified me in recent years has been the way vested interest has muscled into bands (the unlicenced 2.4 & 5MHz bands were supposedly for any sort of access - door bells, CCTV, industirial telemetry, door openers, remote cranes) but wifi has dominted them almost compeltely. Then there were the white space enthusiasts who wanted to use guard bands that had been designed to simplify the design of selectivity of equipment in adjacent channels.
Yes, there is more demand than spectrum. But that is an arguement for letting the regulators take decisions in the public good, not be pushed about by meganationals.
To be fair, I think that Ofcom is doing its best to behave correctly, and it is very careful about prior consultation. But it only regulates these islands, and can be bullied by a steamroller already running from elsewhere. I think the European authorities could have been more proactive.