Because if you consider it from observer dependencies it seem to me that you either have to assume 'something' unchanging, being the platform from where observer dependencies are created, or else make a assumption that all observers have a own 'universe'. And that, it its turn, comes from the fact that we use repeatable experiments to define 'reality'. That's the way we set it up physically. And if that thinking is correct, that your experiments will tell you what is true and what is wrong. Then the universe I see, and experimentally verify as me having a certain distance to some other body for example, will differ from yours according to relativity.

==

Locally I would say that we all have a equivalent arrow, as proved when being in a same frame of reference with what you measure. From a point of locality we're all 'equal'

regarding the arrow. The same goes for distances. From that point of view the universe consist of one unchanging base, same for us all, observer dependencies created through 'c' (combined with energy/mass/'motion'). But the fact is that your reality is defined through your experiment, so finding someone else's watch to go slower than yours relativistically and experimentally must be true to/for you.

So locality is where we're all equal as I see it. And where all arrows are equivalent.

And one more thing, accelerations.

The weirdest example, and proof, of a 'change' that I know. Everything that exist seems to me to somehow (be able too?) accelerate? What is a virtual particle? If it is not there, but then is, can you see that as a 'acceleration' from a probability into a 'outcome' ('real' particle) ?