If I buy a new house in a newer subdivision 5 km from here I can get 1Gbs service.
We have 60 Gbs of external leased lines for our servers now and we would need to buy a lot more if we broadcast 1080p or 4K. Like 10's of thousands of Euros a month ...
Forget about the money -- streaming canned product that is prerecorded is possible assuming people would pay the premium. ATM I am more interested in spending money on new broadcast technologies. It is conceivable that WebRTC peer to peer might support 4K better with the users paying the cost of their own bandwidth. That is preferable to our spending a lot to give away 4K free chat
4K if adopted in webcam streaming will only be for paid private shows to my thinking. It makes sense to deliver the best quality only when people are paying for it. This would limit the costs right now on bandwidth if we use an intermediary server. When you simultaneously transmit 600+ streams to 2,000+- viewers at peak times you would have issues with the FMS servers operating capacity at that speed -- up to 60FPS -- that is 50,000 to 120,000 frames per second transmitted total.
You need to understand loadbalancing, MTU fragmentation bottlenecks and payload delivery in live video streaming. In IPV6
https://en.wikipedia.org/wiki/IPv6_packet#Fragmentation there may be user end fragmentation user CPU reassembly issues too.
Another thing is model (performer) upload speed -- most internet is asynchronous. Studios are sharing fibre bandwidth between their models. Independent models bandwidth will vary also. So it is on the incoming transmission side also.
Experiment toward the future fine. Invest heavily in immature technology -- no.