I’m with you on finding it hard to think they’re just epiphenomena of coupled oscillators. And from the timing perspective and their spatial spread, I think it’s a solid hypothesis they actually contribute to the coupling.

That actually happens in the narrow spatial scale of ephaptic coupling already, where neighboring neurons can sense one another firing and get primed to fire, which helps with synchronization.

Where there’s still debate is in these larger waves, which sweep through very diverse architectures.

And again, part of the answer may be to look at astrocytes, since they can actually tune their syncitium, and some at least can give out glutamate as a signal, and are electrically active: https://www.quantamagazine.org/these-cells-spark-electricity...