There is also another weird asymmetry. For the spin we say that its superposition exists in my [slice of the] world but for the measurement device we say only half of the entangled state, the outcome I become aware of, is part of [my slice of the] world, what justifies this?
Nothing. It keeps existing, and (1/sqrt(2) * <I think spin is down|⊗<device shows spin is down|⊗<spin was down| + 1/sqrt(2) * <I think spin is up|⊗<device shows spin is up|⊗<spin was up|) keeps evolving, and since QM is linear, the summands evolve independently.
Yes, to the outside observer (the one who could look at the whole of the universe from the side/above without interacting with it) this does look very similar to MWI. But we are not outside of the universe, we exist inside it as a part of it, so for us it looks like spontaneous collapse of the wave function.
That idea is almost as old as quantum mechanics itself. (von Neumann, 1932 - if not earlier.)
The "Copenhagen interpretation" as we understand it, in which some thing "collapses" when measured, trades under the same name but was invented later. As a result the original interpretations of Bohr/Heisenberg and their philosophy of science have somewhat exited the discussion, even though they had the only defensible epistemology -- the rest as you say is just ontology.