For my research, we looked at optical vortex modulation schemes for a while. With polarisation (spin angular momentum), there are only two orthogonal states. With orbital angular momentum you can have infinite integer "charge" states that are orthogonal (yes you can have negative charge as well). So in principal by encoding each stream with one of these states, a bunch can be transmitted at the same time in the same beam.
I might venture a guess that this was done over a short distance in a closed room. Problem arrises when these beams counter strong turbulence or scattering, then the vortices are either destroyed or coupled together. So the channel leakage and signal to noise becomes a problem.
Still neat though, bundling 8 channels is impressive.
In a previous discussion here, it was figured out that "twisting" is equivalent to MIMO. Is that your understanding? If so, it will be subject the same capacity limitations, meaning that capacity is approximately linear with degrees of freedom, then tapers off to a limit, the limit being related to the volume occupied by the receiver and transmitter.
Presumably the number of "integer charge" states is in some way related to the volume of space occupied?
Yeah, those problems are part of what makes this so interesting to me, especially since they so are dramatically alleviated in space. I'm betting we'll see this technology used in future space technology, and perhaps future research into the technology will take place up there. The future beckons, and it's partially in space!