Let's hold a few things as true:
Quantum physics requires using probablistic models to describe phenomenom.
Universe means that it contains everything, where by there could only be the one (if something like a multiverse did exist, it would be poorly named, but still within the Single All Containing Thing).
The Universe, from our perspective, is expanding, began with the Big Bang, and will end with the Heat-Death of the Universe.
Within an empty Universe, one might be able to use the same methods describing Q.P. to describe the probability that the entire contents of the Universe would sponteneously come into being.
As an arbitrary point would suffice, a singularity might be used conceptually, bearing an uncanny resemblance to the big bang.
In order to use these concepts, we are forced to seperate time from the rest of physical reality and existing independant of all frames of reference.
Following the HDotU (H.U), all matter (and thanks to E=MC^2 all energy) will be at such a low and dissipated state as to be nare unobservable, yet still expanding as waves or a current, ever outward. Neccessarily it would still command a gravitational effect, even if otherwise conditions became similar to those preceding the BB.
Following all of the above, it seems reasonable to assume that the whole thing could repete any number of times, and thus present the form of QP descriptors in probability, as the same physical laws would govern each itteration, but local results might vary, such that the last itteration in which that photon passed such location it did so just a smidge earlier.
What if, then, Dark Matter is merely historical residue, built-up over the eons, of past manifestations of the physical reality?
If such were true, being able to measure Dark Matter, and then being able to tease apart meaning from it, could indicate, with some degree of probability, how long such a cycle has been going on.
But also, if one could somehow also escape the bounds of time itself, and observe the Universe from beyond time, perhaps we could see all itterations happening simulteneously; but also, as per QP probability, not existing all the same.
If, somehow one could take the long view of this probable dance, it should then appear a dim flickerin light in the distance. And all that ever was, and all that could ever be, might not be known in the same manner you would know an intimate partner, or the topography of a map, but it could be known what is likely true; ultimately that is all that science is really about anyway.
By the way, I have been waiting for this, or similar, thread for a long time