"The paper states that 48 nodes would provide an 800 km quantum link that has a 98 percent probability of successfully transmitting the quantum information. The intermediate nodes in the network could also act as routers."
48 nodes to go 800km. We've already got 60Tbit/s fibres across the Atlantic, and I think those are repeater-less these days. I don't know how many bits they can encode on to each quantum state, but at 10Mqubit/s I suspect they've not actually got a unique selling point, not even security...
"El Reg is aware that perfect "quantum communications" security is a highly qualified and controversial topic."
Good. Let me stoke the flames!
It relies on our understanding (i.e. the theories of quantum mechanics) of a physical effect. If the history of science is anything to go by then the theories are almost certainly incomplete. Who is to say that someone won't come up with a new theory that explains the physics better and simultaneously breaks QC? In short, we vaguely think we know Mother Nature's rules, but we can't be sure.
On the other hand classical cryptography (AES, DES, etc) is based on mathematics, and the nice thing about that is that it is our own invention. We might get the maths wrong and end up building in weaknesses (e.g. Enigma), but at least we wrote the rule book. So at least we stand some chance of being sure one way or the other.