I want to correct and extend a point raised in the thought experiment of the previous post. Under consideration, a blackbody the size and density of the sun placed in the InterGalactic Medium. Such a body should come to equilibrium with the Ambient Cosmic Electromagnetic Radiation. The ACER has never been properly accounted for in its entirety, though a preliminary effort has been made,

The blackbody, being in equilibrium, is emitting as much energy as it is receiving and it therefore has an energy density gradient surrounding it consisting of the outbound radiation which drops off in density as 1/r², the radial distance from the center of the blackbody. The underlying ACER density (presumed approximately constant) does not change with distance and may well be considered an inertial field.

Now we flip the fusion switch and make the blackbody a more realistic astronomical object, a star. Compared to the blackbody, this star has a relatively enormous radiation density gradient consisting of all the omnidirectionally emitted radiation produced by the star. The density of that radiation will again drop off as 1/r².

When a remotely sourced electromagnetic wave passes close by a star the wave is observed to curve as if it were traversing a medium with a density gradient. This is commonly attributed to a gravitational field though such a field has never been observed.

What is observed are a curvature of light and a radiation density gradient. It strains credulity to believe that those two physical facts are uncorrelated. This in turn suggests that observed gravitational effects, commonly attributed to an unobserved gravitational field, are in fact a consequence of matter-electromagnetic energy interactions.