This doesn't look to good nor safe when you look at the numbers. Is it an improvement certainly.. is it safe enough for large scale deployment.. no way.
At best you say you can get 30ms delay at worst 300ms. Far better then the comparable 8+ seconds with other methods. 30ms is triple delay we fly with already LOS on the 2.4ghz band. With 30ms delay using an automated guidance system that delay is super critical to interact with collision avoidance.
Scenerio.. Aircraft is on a planned flight path. Encounters an obstacle. Waiting for control input and getting none the automated measures kick in. the aircraft changes course. Adding in reaction times for the pilot he sees the obstacle and reacts. His signal gets to the aircraft while it is in the middle of its automated routine. It over rides the automated signal and the aircraft reacts to that. Now the automated sees the change and reestablishes control because it is no longer on the point in the path where the pilot intervened. Its like playing ping pong and those counteracting commands will conflict and more then likely put the aircraft into a desync or even failsafe condition.
This does not even cover any situation where an actual piloted method would be necessary. FAR to slow
This is all before even thinking about video transmission delays as I assume your service will only be for command and control NOT any tertiary function like video or telemetry.
Now factor in the special gear to use this (NOT just the onboard stuff) but the network required "expensive" stuff mentioned on your site that providers would have to install and maintain. The cost to use a system like this excruciatingly overrides the cost to travel and actually be on sight for traditional LOS operations or short range BVLOS that is already in place over RF networks.
This MIGHT at some point have an application for commercial use and may be a best solution to a specific case but as a hobbyist.. Yeah.. not gonna happen.