Presumably galactic space is shrinking, aka, gravity.
The reason overall space appears flat is because gravitational contraction and universal expansion balance out. Omega=1. One reason for Inflation is to try explaining this apparant flatness in terms of overall expansion. That it expanded so fast, it only appears flat, like the surface of the planet appears flat to us.
Yet wouldn't it be logical to assume Hubble discovered Einstein's original cosmological constant, the balance to keep gravity from collapsing space to a point? Think in terms of the rubber sheet analogy of gravity, but put the sheet over water, so that where gravity pushes it down, is balanced by the sheet being pushed up elsewhere.
So that when we see light from very distant sources, it has to travel this expanded space. Though if it does pass close to galaxies, is "lensed," aka magnified.
The original assumption was classic doppler shift, with it being an expansion in space, but as it became apparent the overall intergalactic expansion is proportional to distance in all directions, it makes us appear to be at the center of the universe. Which was when the theory was changed to say that space itself expands.
If space expands, it still has to expand locally in order to expand everywhere, so it expands where the light/photon is traveling. If this doesn't increase the speed of light, so that it remains constant, then two metrics of space/distance are being assumed and since the evidence for this expansion is the spectrum of that intergalactic light, it seems both are derived from the same light.
Given these metrics are based on the same light, thus related, but different, than which is the denominator and which is the numerator? If lightspeed were the numerator, than we would be discussing some form of "tired light" theory, so it seems the speed is still assumed to be the denominator, the actual "ruler," so either we are at the center of the universe, or, given we are at the center of our point of view, cosmic redshift is an optical effect.
While single spectrum light will only redshift due to recession, multispectrum light "packets" will redshift over distance, as the higher frequencies dissipate faster;
2008CChristov_WaveMotion_45_154_EvolutionWavePackets.pdf
The Pandora's box this opens is that we would be sampling a wave front, not seeing individual photons that have traveled billions of lightyears. That throws a monkeywrench into the particle side of the wave/particle duality.
I would note though, that waves tend to either synchronize, which is centripetal, or harmonize, which is centrifugal. So they tend to form nodes and networks. Organisms and ecosystems.
Particles and fields.
Black holes and black body radiation.
Consider that quasars shoot out the poles of black holes and quasars are ginormous lasers, which are synchronized lightwaves.
The amplitude loses energy, the frequency is the problem.