Creating Energy and Counter Speculations

Ok back to time.
My favorite astounding quote of the week

"Now in cosmology, energy is not conserved", P.C.W. Davies, 1983

Ok, so is energy conservation just a tool for doing simple physics that we naive youngins cherish blindly like some kind of constitution or Bible? The way he throws that out there bothers me almost as much as the fact that he says it all. Apparently this is the nature of the argument behind inflation. Some energy must have been created to power inflation. This energy enters in the form of "Dark Energy" (or "Dark Tension" or a positive cosmological constant or an equation of state parameter w<0). is not invarient under time translation because processes are not reversible. Does it then follow that energy is not necessarily conserved? Is the energy-time proof true in the negative sense? Is there some way to translate the question from time to entropy? So that I can find a new conserved quantity or perhaps a dynamical equation for the total energy?
...
Reading the Penrose 1979 and then immediately Penrose 1989 is very interesting. In 79 he made the argument that collapsing the wavefunction was a time reversable process then in 89 he argues the reverse. From my perspective I have to say that I never really believed that it was a time reversable process. Thats kind of the whole point isnt it? That once the wavefunction is collapsed and an observation is made then that particular aspect of the system behaves in a classical way.
For example, consider a two slit experiment. Let a beam of electrons pass through the two slits and one sees a diffraction pattern, decrease the particle rate to one and you can still predict that the particle is not very likely to hit at a null in the diffraction pattern. You can build a diffraction pattern one particle at a time, thus the particle is behaving like a wave. Now put a device at the slits which triggers when a particle passes through, perhaps by some E-field measurement. Now you dont get a diffraction pattern, not exactly sure what you do get though. Probably just one spot for each slit. In the time reverse experiment you evolve two beams of electrons backwards you measure the position as the particle passes through and you get two beams coming out definitely not the same.
The other thing I am thinking about recently is the initial correlations Page (Nature 1983) talks about as the assumption hidden in the "inflation exlains time asymmetry" argument. He says that assuming that each point in space was uncorrelated at the "beginning" assumes a much smaller parameter space than could be. (Which to me is another way of saying Penrose's "Creator with a pin in phase space" argument. (Penrose 1987) I guess parameter space is probably just a more general phase space. So I guess the way one goes about showing uncorrelation is by formulating the size of a points past light cone and the size of the universe in terms of the same parameter (time?) and then take the limit (t->0) and ask the question " "which goes to zero faster?" My guess is that maybe you get some space with a mathematical property which something like an infinite set of unconneted singularities. I guess I picture a spate-time foam prior to the planck time, take each on of those foam cells and contract it even further at the foam walls the light cones touch, as they contract further they seperate so they are no longer causal with each other and thus not connected.