>The twin scenario was presented by Langevin in 1911 to show that >physical acceleration is "absolute", even more so with SRT than with >Newton's mechanics.
What does that mean? As I said, proper acceleration (as measured by an accelerometer) is absolute, but coordinate acceleration is certainly not.
>He argued that these absolute effects detect the ether (what you call >a "preferred frame").
If that's what he argued, then he was wrong. The fact that acceleration is measurable does not imply the existence of a preferred rest frame.
Here's an analogy: A flat Euclidean plane has no notion of a preferred direction. Any direction is as good as any other. But it certainly has a notion of a *change* of direction. If you draw a path on the Euclidean plane, then you can unambiguously determine whether the line is straight or curved, because a straight line connecting two points is shorter than any curved line connecting the same two points. If you measure the lengths of two curves, you can determine which one is straight.
A rest frame in Einstein's spacetime is analogous to a direction in Euclidean space. There is no preferred rest frame in spacetime any more than is a preferred direction in the Euclidean plane. But a *change* of rest frames is certainly detectable, in the same way that a change in direction is detectable in the Euclidean plane.
>However, Einstein (1916) considered that the PoR of SRT has an >"epistemological defect", since it relates to a privileged group of >"spaces" that cannot be observed. And what he could not observe, he >called 'factitious'. In other words, he rediscovered Newtons' argument >but he found it unacceptable. He preferred to go the opposite route >and extended the PoR as follows: > >"The laws of physics must be of such a nature that they apply to >systems of reference in any kind of motion".
>As a result, physical acceleration is, according to Einstein's GRT, >*relative* - which is just the contrary of what Langevin argued based >on his "twins" example of SRT.
As I said, proper acceleration is definitely *not* relative, but coordinate acceleration trivially *is*. But proper acceleration is measuring acceleration relative to *freefall*.
>It should not be surprising that this was not only very confusing for >bystanders (who already hardly understood the difference between the >two theories), but that it even looked like a contradiction
I would like to hear any coherent explanation of why it looks like a contradiction. The bare statement "The laws of physics must be of such a nature that they apply to systems of reference in any kind of motion" is not a contradiction---on the contrary, it is nearly a tautology. You can always write the laws of physics so that you can use an arbitrary coordinate system.
To derive a paradox from the twin thought experiment, you need to reason something like this:
1. There exists two coordinate systems, C1 and C2, such that the path of the traveling twin, as described in C1, is the same as the path of the stay-at-home twin, as described in C2.
2. Therefore, the predicted age of the traveling twin, computed using C1, must be the same as the predicted age of the stay-at-home twin, computed using C2.
I don't see how 2 follows from the general principle of relativity, as expressed in the sentence "The laws of physics must be of such a nature, blah, blah, blah." From the latter, it follows that one can use either C1 or C2 to compute the ages of the two twins, but it *doesn't* imply that the ages will be the same. To be able to conclude that, you need to assume a very specific form for the laws of physics.