Until recently, I was under the impression that the invariance of light was an experimentally proven fact. But it turns out that all such experiments testing for this (e.g. the Michelson-Morley one) actually only measure two-way light-speed averages, and not precise one-way ones. From there, it follows that there is an essentially infinite class of anisotropic light-speed theories (as developed in the vein of the work of Mansouri & Sexl) all of which are perfectly capable of reproducing the exact same mathematical predictions of Special Relativity, but none of which feature isotropic/invariant light. I then further learned that Einstein himself wrote in a 1908 paper that his theory and Lorentz's Theory, the latter of which featured absolute space/time and an ether, could not be experimentally distinguished.
My question is this: doesn't the One-Way Speed of Light Problem immediately invalidate the Einsteinian interpretation of Relativity? That is, we're taught that phenomena like time dilation, length contraction, or relativistic mass only exist as consequences of the invariance of light. But every anisotropic theory also features time dilation/length contraction/etc. -- but no invariant light. So how can invariant light "explain" these phenomena, if the same phenomena are likewise derivable from anisotropic light assumptions in other non-standard formulations?
Moreover, doesn't the one-way speed of light problem throw the whole conception of "frame of reference" into some seriously questionable territory? I was under the impression that a frame of reference was something physical, like a system of co-moving rods and clocks, to which the motion of light could be relative. But it's clear from the one-way speed of light problem that to set up a frame of reference, one first has to make an assumption about the behavior of light, i.e. isotropic or not, in order to establish a notion of simultaneity and begin assigning coordinates. This means the lengths of rods and time durations of clocks are not set until the observer chooses a synchronization convention (Einsteinian or otherwise). This makes a frame of reference an arbitrary choice of convention, not a physically real thing. So how can the motion of light be relative to it?
That is, if we consider the following three assertions as being true:
A) Light is a physically real thing (i.e. exists outside and independent of human perception/awareness of it).
B) The motion of physical entities must be relative to other physical entities.
C) A Frame of Reference is not a physical entity, since it depends on an arbitrary choice of synchronization convention.
then I don't see how we can accept the Einstein interpretation of relativity, since it requires the motion of light be relative to a reference frame, which produces a logical contradiction.
So what's going on here? Who can explain to me why these apparent contradictions seem to go completely unaddressed?
