One of the biggest challenges in distributing entanglement is loss. It's easy to send a classical signal over 30 km of fibre. You just send a lot of photons a

One of the biggest challenges in distributing entanglement is loss. It's easy to send a classical signal over 30 km of fibre. You just send a lot of photons and a significant portion of them will get through. Big pulse goes in, smaller pulse comes out. All you care about is that there was a pulse. When you're trying to distribute a pair of entangled photons, it really is important that you're sending one photon to point A and one photon to point B. One photon goes in, and maybe it comes out. Maybe. If one half of a pair gets where it's going, then you need to be lucky for the other half to get where it's going too, otherwise you have to throw the one that got through out. Typical quantum entanglement experiments have multiple massive sources of loss. Need to use a 50/50 beam-splitter somewhere? There goes half your photons. Any given mirror incurs some loss. Detectors may only be able to detect a tiny portion of the photons that hit them. The list goes on and on. The experiment here was done using typical telecom fibre. Although telcom fibre is "good" at transmitting certain wavelengths of light, we're still talking about shining light through 30 km of glass here. That's a lot of loss. Think about looking through a window that's 30 km thick. A free-space link with a satellite is going to have significantly less loss. First off, the air you're breathing right now is a lot less optically dense than glass. Go up 30,000 km, and it's even less dense. By the time you get to the satellite, it's vacuum. In total, a free-space satellite link incurs a lot less loss than a fibre link this long. The distance of this experiment is indeed far shorter than the freespace one, but they're dealing with much higher loss. The motivation for this experiment was practical quantum networks using existing and installed fibre. These are both impressive experiments and you're right that "distance" should be qualified a bit here, but free-space and telecomm fibre experiments really are apples and oranges. Free-space experiments have their own challenges, and free-space links have their own applications. If you want to build a QKD network that's for commercial use, a fibre based one is probably a lot more practical.