It’s possible that NREM3 (when the glial network is most active) plays into this. There’s some evidence that astrocytes (a kind of glial cell), although not electrically active , can mediate synaptic response. There’s an overview of current work here.

NREM3 can be induced by transcranial electrical stimulation, but it seems most work with this has been done with differenntial memory, not figuring things out — so maybe something to research?

Anecdata: If I force myself to read something I don’t 100% understand just befor I go to sleep, I often find myself explaining it to people the next day.