To put this into context, retinal waves (essentially self-propagating excitations along the retina) pre-birth have been known to exist and help neural organizat

To put this into context, retinal waves (essentially self-propagating excitations along the retina) pre-birth have been known to exist and help neural organization for what must be 20-30 years. What's new in this paper is that they additionally confirmed that the direction of the propagating waves coincided with the mouse's typical optical flow pattern, essentially helping priming motion-detection neurons in addition to the well-known edge-detection neurons early. I wouldn't compare this to dreaming really, it's a very different process where top-down connections hallucinate sensations. This is a completely bottom-up process, it's more or less like a test-bench connected to the optic nerve generating moving edges pre-birth.