AFAIK, the force of gravity propagates at light-speed, so every particle in the universe is not being acted upon by every other particle in the universe via gravity, because the distance between some particles is increasing faster than that due to the expansion of the universe.
All that means is that movements of particle A and thus changes in its gravitational field now will never reach particle B, which is so far away that the relative expansion rate is c or more. However the historical gravitational field of Particle A, the part of its field that is within particle B's light cone, will always affect particle B.
What if the light cones of A and B don't intersect at all? There are certainly places in the universe so far apart that light not only can't now, but could never, have travelled from one to the other; that is part of why the near-isotropy of background radiation is something that needs to be explained, I thought.
Are there regions of the universe that have never been causally connected? They all originated in the same event, so I'd be surprised if that's true, but I'm no expert. I know cosmic inflation during the big bang was superluminal, but I have to concede I'm not completely familiar with the consequences of that, so good point.
Indeed, the fact that they are moving apart at superluminal speeds is exactly why they have never been causally connected. The fact that such non-causally connected islands still "look the same" is called the horizon problem (https://en.wikipedia.org/wiki/Horizon_problem), and it is one of the issues that Guth's inflation might solve (https://en.wikipedia.org/wiki/Inflation_(cosmology)#Horizon_...).
I think the reason it doesn't count as a measurement is that the the gravitational effects are so weak the no-far-off-particle and yes-far-off-particle outcomes overlap substantially.
You start with a cross-section of where a "measurer" particle might be, and if the far-off "to-be-measured" particle is present its gravity shifts the cross-section of the measurer particle's position very slightly, but there's still a lot of overlap with the original cross-section. A proper measurement would have nearly no overlap, instead of almost complete overlap.
Every particle in the universe is acted upon by every other particle, via gravity, so does this not mean that every particle, in some small capacity, "observes" every other particle?