Until you have a working model that can correctly and precisely predict these phenomena, I'd argue that they constitute evidence against whatever you're proposing rather than evidence for it, when GR with a few tweaks already works almost perfectly.
I don't fully understand what you're trying to refer to with this concept of "energy equalization", because you haven't hashed it out in any significant detail. At face value it doesn't sound at all like any feature emergent from the GR picture, and you can't just toss whatever features you want into a mathematical model and expect it to remain self-consistent. So by my understanding of what you're proposing, adding this concept of "energy equalization" to the picture means all the GR equations would have to be tweaked or completely replaced in some fashion to allow for it.
All the same, if the galaxies are clustering up at the local level, why would energy density shift over towards a uniform distribution on global scales? A priori there's no reason to contemplate such a picture, it doesn't emerge from General Relativity as it's presently understood, and in terms of physical evidence, the universe is already statistically homogeneous and isotropic, which means there's no varying energy densities to shift around anyhow, on global smeared out scales it's already uniform as is.
Well, the best I could personally do with the time and knowledge available to me was to picture a finite bubble universe operating under Newtonian gravity (which is still legitimate as a weak-field approximation to GR) with energy densities varying at different radii from the center, and there are several reasons why it doesn't appear to have any chance of working as far as yielding the features you're looking for, let alone predicting redshift, CMB and all the rest. To apply General Relativity in its full form directly to this picture would be exceedingly difficult and well beyond my present capabilities; it's a nightmare to work with even in the simplest possible cases of uniform energy density or eternal universes containing nothing but a single spherically symmetric black hole, and to go beyond that you basically have to resort to computer modelling and numerically approximating solutions which frequently don't differ much from the Newtonian prediction.
My current understanding of topology isn't sufficient to tell you what would happen in a closed spacetime where there's a limit to how far space itself extends, but in all honesty it doesn't seem to me like any of your ideas are particularly revolutionary things that no theoretician would have already tried to play with, and if placing some kind of boundary on spacetime to make it loop back on itself with varying energy densities and all the rest looked like a promising approach, I'm pretty convinced it would have already been attempted and published. Well, I did what I could with the tools at my disposal, and if you think there's something important I missed, I'm happy to take another look. I think exercises like this one right here are valuable for showing why scientists don't put much time or stock into certain concepts, but beyond direct modelling attempts I don't think there's any scientific legitimacy to qualitative verbal speculation any more than you could announce the sum of a series before you've made any attempt to calculate it.
Edit: So let's please stick to what we can mathematically model to one degree or another and compare directly with real evidence, and leave the "maybe this, maybe that" preaching to Philosophy, Alternative Theories etc. where it belongs.