Perhaps I need to state the case more concisly.

Postulate
The Hubble expansion is an expansion of the spactime interval.

It is assumed that if the Hubble expansion is an expansion of space then it must be reflected in the spacetime interval s² = x² - c² t² of the observers in spacetime. This requires that the Hubble expansion is not observable locally for the same reason that an observer cannot observe ∆s locally as the depth in a gravitational field changes. Such changes are only observable by comparing ∆s to another reference observer as the depth in a gravitional field changes. We can do this by measuring the gravitational redshift of light sent from one gravitational potential to another. To observe the Hubble expansion we can measure this same redshift by recieving a light signal from the past.

The gravitational redshift is z=(ω_o-ω_e)/ω_e.
ω_o is the wavelength as measured by a distant observer.
ω_e is wavelength as measured at the source of emission.

The Hubble redshift is then z_h=(ω_o-ω_e)/ω_e.
ω_o is the wavelength as measured by a future observer.
ω_e is wavelength as measured at the time of emission.

Nothing more than the postulate and the standard model of physics is needed to make the case.