No offence, but Sol doesn't know as much as he thinks.
That is probably true. Nevertheless, your posts make it plain that I know far more about this than you do.
I reiterate: if you assume that the universe is homogeneous and isotropic, there's absolutely no gravity in it, spacetime is flat.
That is flat-out false, as W.D.Clinger and others pointed out. To see that it is false, all one needs to do is compute the curvature - which for FRW solutions is extremely easy - and note that it is not zero.
Lay it on me, sol. But do note that understanding the simple case relates to axioms, essentially undertstanding the terms. That's sometimes an issue.
Sure. I have a deep understanding of say
this that gets right down to the fundamentals.
Good. Very well. Start from Newtonian gravity, because general relativity reduces to Newton in the limit of weak gravity (as any potentially valid theory of gravity must). Here's Newtonian gravity:
[latex]$\vec \nabla \cdot \vec g = -4 \pi G_N \rho_m$[/latex], where rho_m is the matter density, [latex]$\vec g = \vec \nabla \Phi_N$[/latex] is the gravitational field, and G_N is Newton's constant.
Note that if you map the gravity field to the electric field and mass to charge, this is identical to Gauss' law in electrodynamics, which the exception of the minus sign (which means positive masses attract each other instead of repelling).
Now, we know gravity should act on all forms of energy, including the energy in the gravitational field itself. So if we want to include the effect you mentioned, we should start by adding a term to the right hand side that encodes the energy in the gravity field. But what term should we add? Well, first let's recall what the energy density is in an electric field:
[latex]$\rho_E = (1/2)\vec E^2 = (1/2) (\vec \nabla \phi)^2$[/latex],
where phi is the electric potential and E is the field.
Question: what is the energy in an gravitational field? The obvious guess is that it should be given by this same formula, namely [latex]$\rho_g = (\alpha/2)\vec g^2 = (\alpha/2) (\vec \nabla \Phi_N)^2$[/latex], where alpha is some constant. So, let's see what happens if we add this term to the right hand side of Newton's equation:
[latex]$\vec \nabla \cdot \vec g = -4 \pi G_N \rho_m + c^2(\vec g)^2/2$[/latex], where c^2 must be there by dimensional analysis (the fact that it's c^2/2, and not c^2/53 pi or something cannot be derived from this, but if you're more careful and check the constant I believe that's exactly what you find).
A couple of things to notice about this formula.
First, since the second term on the right is quadratic in g, it's not important when g is small. That had to be the case since the theory needs to reduce to Newton in the weak field limit.
Second, if the field g is weak, we can easily estimate how important the second term is. All we need to do is calculate what g would be in Newtonian gravity, plug that into the second term, and compare it to the G_N rho_m term to see how much it matters.
Before doing that for a galaxy or other interesting object, I'll pause and make sure you're following so far.