Eckard
""everything is moving" ??? If the universe is moving, what does its speed refer to?"
No, I am saying we must presume everything is moving, which effectively is the same end result, but the correct expression. Because we have no absolute reference against which to make a judgement. And if you, or someone else you know thinks otherwise, explain to me what that something is, and why you know it is absolutely stationary. In other words, when we choose a reference, which could be anything (but it has to be something) then that is conceptually deemed to be stationary whilst all other speeds/directions are calibrated against it, ie it either appears to be stationary, or if we can see that it is relatively moving we can discount the effect over time whilst measuring (aka translation).
And presuming that the 'universe' is another label for 'all of existence that we can know', then the answer is that we do not know anything extrinsic to that. So it is irrelevant. We can only have knowledge. When we think we are comparing knowledge with reality, we are not. We are comparing knowledge with other knowledge. We can only know, and we can only know what it is possible for us to know. Which, by definition, does not include what we cannot know. We can think up a countless number of ideas, but that is not knowledge/ science.
Now, if your 'universe' relates to a sub-set of all we can know, then we just need a reference extrinsic to 'universe' to calibrate its speed. Which might well be difficult, but that is the point.
Another point to make here is that in either case, but especially the former, this effect is omnipresent, so it is irrelevant for calibration purposes, which is about comparison to identify difference. Discounting an effect that is everywhere makes has no effect on difference.
"I consider the propagation of a wave related neither to its emitter nor to any receiver but to the medium of propagation."
Fine, that is as good a choice as any for the reference against which to calibrate. Now tell me what specifically it is, and how you identify, and then maintain identification, of a specific entity of this 'medium' in order to effect all the calibrated measurements. Remember, to ensure comparability of measurements, the same reference must be used (or adjustments factored in to effectively make it the same), otherwise they are useless. Because I can say that A measured X speed and B measured Y speed. To which you would think there was a difference, but then get annoyed when I reveal that A was moving X when measured against D, whilst B was moving Y when measured against F. You would immediately point out to me that I have to use the same reference, or what was the difference between D & F, which of course requires another common reference to establish. This is the mistake Peter, and others, keep making in their emission theories, the reference is changed to maintain the calibrated speed.
So you have still not addressed my point that light vis a vis earth calibrated speeds must vary depending on how the two are compared.
"And Michelson's - as I maintain wrong - expectation eventually led to SR"
As explained, again, above, presuming what expectation you are talking about (ie light/earth speed), that was not wrong. Whether they came to it for the wrong reasons is another matter. Neither did it lead, in any sense of that word, to SR. Because SR is a theoretical circumstance where there is no gravitation and everything is fixed shape and moving relatively constantly. Neither did it 'lead' to relativity. What 'lead', in the sense of created the mental environment which resulted in the construction of this particular perspective was a certain presumption which was fulfilled by invoking errors. The flaw in relativity is in those errors. As I keep on saying, and have explained elsewhere, the problem in relativity is not the constancy of observational light. It obviously cannot be, because it is more or less constant, so such a simplification cannot invoke a fundamental flaw.
Paul