I'm sure my comments above are pretty opaque, so let me try to give a little more perspective.
The conventional way of looking at things is Newtonian - the universe as a static eternal back drop. And this stage has atomistic contents - a bunch of particles.
Modern cosmology challenges this basic picture in so many ways. And a new core mental picture could be emerging - at least you can see the gist of things in the writings of Davies and Lineweaver. I've read a ton of different people, but I keep coming back to their current work as lighting the path ahead.
See
http://www.ctnsstars.org/conferences/papers/The%20physics%20of%20downward%20causation.pdf
http://www.mso.anu.edu.au/~charley/papers/DavisDaviesLineweaver.pdf
and really many of the papers on
http://www.mso.anu.edu.au/~charley/publications.html
In brief, they are taking a self-organising system view. They are taking the observed facts about the universe, such as its scale flatness, its expansion, its light cone coherence, and making them basic rather than add-ons.
The Newtonian universe is static, so expansion and a direction of change have to be added as secondary features. Likewise it is empty, so contents like energy have to be added.
This is all very primitive modelling, and so we want to start with a model that is inherently dynamic, inherently full of potential (potential energy and potential spacetime), etc.
Now the beautiful idea at the heart of the Davies/Lineweaver work (it is not their idea as such, many are thinking along the same general lines, they just express it most clearly IMHO) is that the description of the universe can be boiled down to blackbody photons radiated by event horizons. A de sitter space view.
It is a spherical co-ordinates type view (and I remember nismaratwork may have had an enthusiasm for that). Instead of spacetime as a static frame, it is defined in terms of lightcones - comoving volumes bounded by event horizons. So an inherently dynamic view. And event horizons have now been brought into the fold of thermodynamic/entropic modelling. So the boundaries of spacetime are inherently in dynamic equilibrium with the contents. Energy is being related to spacetime in a direct fashion. Closure, in energy conservation terms, is being defined (in a way it was not in an open, unbounded, Newtonian universe).
The event horizon approach also unites the description of the local and the global. The global scale is set by lightcones and comoving volumes. The local by the thermodynamics of black holes.
We know the universe is expanding, which is puzzling because it seems like an action in a single direction, a violation of conservation. Some force must be continually pushing it larger.
And then we look at the contents of the universe, such as photons, and say something must be stretching them - like the expansion of space.
But an inherently dynamic view of the universe puts these kinds of things back together as a single action of separation - symmetry breaking. Energy spreads out to expand spacetime, and the growth of spacetime cools energy (spreads it about). Two faces of the same coin.
Again, you then need an ontology of development that makes this an actual change in state, rather than an apparent tautology, a simple circularity. It certainly sounds circular to say that photons are spreading the spacetime they are in, and spacetime is spreading the photons it contains. Which is why a notion like vagueness becomes essential because you do get back actual change - a change from a vague potential to a crisply dichotomised outcome.
But regardless, current cosmology - taking Davies and Lineweaver as examples - is learning to take a dynamic and self-organising approach to imagining the universe. It is quite unlike the Newtonian thinking that is still the standard issue mental model.