What I am saying is that theta waves would ordinarily be blocked from a part of the brain when a person is awake. That part of the brain is responsible for sleep paralysis.

Theta waves are also involved in the short term memory event. They probably increase in magnitude during the short term memory event. (you'd have to google some brain wave charts to see the fluctuations)

Since the channel to the sleep mechanism is open or partially open the theta event registers in the sleep part of the brain. Given enough events registering (say 40 events), the sleep mechanism switches on fully.

Eventually the mechanism looses it's effectiveness and subsequent theta impulses receive an impotent response but the events continue to interfere with the normal functioning of the brain.

Sometimes the theta event that is passed to the sleep mechanism is rejected. This results in an anti-sleep response due to the perceived error. The anti-sleep response invokes a rejection of the present state. Depends the lifestyle and environment. Non-rejections types are probably more meditative where as rejection types are more structured.

This is a model which is very distinct from the concept of "brain stimulation" or an overall lack of neurochemicals. For that reason it rejects the concept of partially medicating or taking medication breaks. Most of the sedation or other concurrent symptoms come from the neurochemical release of the sleep mechanism. Once the pathway is blocked there is no need for further stimulation. It is only when the medication is not effective enough to block the channel that there are manifestations of symptoms.