On first reading I'm not entirely sure if there's any great theoretical breakthrough here (... if someone can distinguish the innovative parts better, go right ahead, I'd appreciate that), compared to what banks have operated as "secure" data links for a while now. The technology seems vastly more portable though, the network structure something like what many hub centric pre-internet network ideas were. I think I've seen other kinds of quantum switch ideas out there too.
It's debatable how "uneavesdroppable" quantum comms are, some researchers have come quite close to levels of interference that wouldn't raise alarm with Alice or Bob (... I don't think these devices are anywhere near down to a bit per a photon rates). There's also quite a chasm yet to be bridged to integrate this with current quantum computing capabilities such as DWave, quantum walk devices and so on. But overall this is a good problem, as we're further along in this area than most assume.