This is exactly backwards. Einstein's epiphany wasn't that gravitating objects cause nearby objects to accelerate, or that gravitating objects cause nearby photons to accelerate as much as matter does. Einstein's epiphany was the opposite of that, namely that gravitating objects do not cause nearby objects to accelerate!
An object in free fall doesn't experience any proper acceleration. A free-falling object only appears to accelerate, if you use a non-inertial frame of reference. That is, if you choose to use a non-inertial frame of reference, gravity will appear in the form of a purely fictitious force.
In a version of Einstein's "elevator" thought experiment, free-falling objects (or light) inside an accelerating rocket in space undergo no proper acceleration. However, the free-falling objects mathematically behave similar to accelerating objects, if for convenience reasons you choose to describe the objects using an accelerating frame of reference in which the rocket's spatial coordinates don't change with time. Completely equivalently, free-falling objects inside the rocket undergo no proper acceleration if the rocket is sitting on its launch pad on Earth. They merely mathematically behave similarly to accelerating objects, if you choose to describe the objects using a frame of reference in which the spatial coordinates of the rocket, and the Earth's surface that the rocket is sitting on, don't change with time. A coordinate system "attached" to the Earth's surface is not an inertial frame of reference, it's an accelerating frame of reference whose origin accelerates "upward" (away from the Earth's center).
From the perspective of gravity as a fictitious force, of course photons will appear to accelerate in an accelerating frame of reference, with the same "acceleration" value as a bit of matter, because the "acceleration" has nothing whatsoever to do with any consideration of whether or not photons interact with a "gravitational field". The apparent acceleration is entirely due to choosing to use a non-inertial coordinate system.
One can't really be blamed for using a non-inertial coordinate system near gravitating objects, because it isn't possible to use an inertial frame of reference that covers all of a region of spacetime around the gravitating object. A global inertial frame of reference like that does not exist, due to the gravitation object causing curvature in the spacetime around it. (You can, however, use inertial frames locally near an event, or approximate all of spacetime under appropriate conditions as involving a small perturbation tensor field on a flat spacetime background.)
Gravitating objects not causing nearby object to accelerate isn't simply a matter of perspective. In general relativity, there's a physical difference between using an accelerating frame of reference, and modeling the situation with a proper force that accelerates everything. Using an accelerating frame of reference results in gravitational time dilation, which has indeed been observed.