J O Linton said:
The wavelengths don't expand any more than objects traveling at near the speed of light physically contract.
No. As
@Vanadium 50 has pointed out (and as John Bell stressed in his paper "How to Teach Special Relativity", in which he introduced the Bell Spaceship Paradox), length contraction is real--it's not just a matter of "perspective". If you arrange your experiment right, a length contracted object can fit through an opening that it would not fit through if it were at rest relative to your apparatus. That means length contraction is a real, physical phenomenon.
Similarly, the difference between the "cosmological redshift" case and the "light trapped in a box" case is a real, physical difference, not just a matter of perspective. See further comments below.
J O Linton said:
The truth is that, even though the galaxy which emits the light and the receiver on Earth are not moving with respect to each other (they are both 'comoving' observers)
Wrong. The emitting galaxy and Earth
are moving with respect to each other. The fact that they are called "comoving" observers does not change this; "comoving" is a technical term that means both observers see the universe as homogeneous and isotropic (actually, here on Earth
don't exactly see that, since we observe a dipole anisotropy in the CMB that we have to remove before analyzing the data in order to draw conclusions about "comoving" observers in our vicinity, but we'll ignore that technical complication here).
J O Linton said:
they do not share the same inertial frame. In the galaxies frame when the light was emitted it had a wavelength of 600 nm but in the Earth frame, 10 billion years later when the universe has expanded by a factor of two, the same wave appears to have a wavelength of 1200 nm.
This difference has nothing to do with frames; you can do the computation of the emitted and received wavelengths in any frame you like and you will get the same answer. The difference in wavelengths is due, as I said before, to the relationship in spacetime between the emitter, the light, and the receiver. That relationship is invariant, independent of any choice of frame.
J O Linton said:
On the other hand, the waves in the box stay in the same inertial frame the whole time because they are confined within a bound system. Their wavelengths therefore do not change.
Again, this has nothing to do with frames; you can calculate the wavelength of the light using the global "comoving" frame for the universe and you will get the same answer. The wavelength being unchanging is, again, due to the relationship in spacetime between the light and the walls of the box (which play the role of "emitter" and "receiver" in this scenario), and that relationship is invariant, independent of any choice of frame.