Does the CMB give us absolute space and time in practice?

[vanhees71]

It's a frame that corresponds to a set of symmetries (Killing vector fields) of the spacetime, and therefore makes the metric look simpler, yes. But that's not what "preferred frame" means in the physics literature; that term refers to a frame that is picked out by the laws of physics, not by particular symmetries of a particular solution to those laws. And of course examples of frames that are picked out by particular symmetries of a particular solution are ubiquitous in GR; the "CMB frame" is by no means the only one. That is why multiple posters have objected to using terms like "preferred frame" or "absolute space and time" to refer to FRW coordinates (which is what the "CMB frame" amounts to).

I said that repeatedly it's not preferred by the general laws of nature but by the physical situation considered (large-scale coarse-grained model for the Universe). There's of course no absolute space or time in the foundations, but the physical situation distinguishes the local rest frame of the CMBR as particularly "natural" to describe this situation. That's also the reason, why by definition (or if you wish convention) intrinsic properties of a medium (starting from elementary particles with their mass and spin; then temperature, chemical potential, particle density, various thermodynamic potentials to characterize thermal equilibrium of a medium) are defined in the (local) restframe of the medium. It's just a matter of convenience to define such quantities in a "preferred" (local) frame of reference and then formulate them in a manifestly covariant way to make them easily treatable in general frames.