If what I claim is correct, then we would expect the theory to explain many aspects of mental life and perception, including dreams. Consider our distant ancestor, a land based exothermic vertebrate with a median eye on top of its skull. During the day there is sun to warm its body and it moves around seeking food and avoiding predators. At night when it has become cooler the animal is unable to move around and becomes stationary. The median eye cannot be closed and continues to survey the feint measures of illumination received from the moon and stars.
Despite hugely reduced metabolic activity, occasional ‘twitches’ might occur, triggered by fluctuations in moonlight caused by passing clouds. The creature would have experienced a reduced intensity ‘simulation’ of daytime activity, the twitching being a scaled down day-time muscular reaction to shifts in light. It is worth noting that whilst modern mammals (and to a lesser extent, birds) do seem to dream, their common ancestor the reptile does not. When humans dream, I suggest the ‘mind’ that cannot be switched off continues to register a ‘shadow’ or ‘phantasy life’ which resembles or simulates waking life, but with locomotive and sensory systems ‘disengaged’ allowing greater freedom for subjective consciousness. This phenomenon is highly comparable with the night-time ‘median vision’ of our exothermic progenitor ... the difference being that our dreaming is ‘symbolic’ in content whereas the median vision of the stem-reptile was ‘direct’.
There is a more technical level to MVT of origins of dreaming. The Academic Papers on origins of REM, phasic transients, loss of external clock and resultant change of brain circuitry from (hardwired) finite-state to a (softwired) ANALOG-ous to the warm-blooded brain with infinite-state circuit morphology are now available: