David Carroll writes:
And then wouldn't this account for the red-shift? If c were faster, then wouldn't the space dimension parallel to the direction of the faster-c light-beam undergo a Lorentzian expansion/dilation?
Perhaps. But a larger, simpler problem is created.
If "old light" is grandfathered to be faster-c, and "new light" is a slower-c... then we should be able to measure a difference between the c on new light from a lightbulb turned on here on Earth vs. the c on old light coming into Earth from far away stars.
But... I don't think a difference is measured.
In fact, I think that the two speeds of light (and any other attribute of the new vs. old light other than age) are exactly the same.