When the rubber meets the road the best ways of working stuff out is to shut up and calculate, you don't figure out anything by assuming unobservable fundamental indeterminism.

No one assume it.

ok, postulate

No one does that either. The nondeterminism in QM is basically just an experimental fact which sort of inveigles itself into the structure of the mathematical framework which makes it hard to ignore when one tries to understand what the theory may mean philosophically.

I've been to phil. physics conferences and I've never encountered anyone who has any kind of strong attachment to nondeterminism. In fact, in general, I think almost every physicist who learns QM has a prejudice against it which never entirely goes away unless they get deep into foundations which forces a more detached perspective.

Not really, "it just is" isn't how facts work. Preferably it should be a quantity.

But it's the opposite, all people with shallow understanding of QM believe in indeterminism, because it's the first thing they learn from the start, and then this belief ossifies. Oh and Heisenberg uncertainty principle tells you that reality is unreal. Conversely foundations are deeply mathematical with basically nothing uncertain or random is sight. In fact it's foundations that have bias for mathematics and determinism. How do you interpret the evolution operator as random?

I don't really get your point - the time evolution operator is deterministic, but it describes the wave function, which itself gives a probability distribution for the outcome of measurements. Typically exegeses of QM are roughly given in two parts: deterministic time evolution of the wave function and random selection of the outcome of measurements. So it really is just right there in the formalism: some stuff is random.

From the point of view of foundations of physics there really is no easy way out of this apparently fundamental randomness. We can be wave-function substantivalists and then we must explain why and how we don't see wave functions but concrete outcomes. We can adopt t'Hooft's cellular automata interpretation but that framework cannot easily support even something as simple as basic interactions in QFT (last I checked). We may list any number of ways of interpreting QM but all of them that I know of only at most banish randomness to the initial conditions, but not totally eliminate it from the ontology.

I'm not saying that reality is fundamentally non-deterministic. But I am saying that most of the ways one squares basic quantum mechanical predictions with basic physical measurements suggest a type of uncertainty that is at least very close to being "fundamental."

Evolution operator shows that state evolves deterministically and doesn't hide randomness anywhere. You use probabilities when you want to explain human experience, because you can't practically compute human's state. But you don't strictly need fundamental randomness for this optimization, Bayesian probabilities work in the same way, whether randomness is fundamental or not is difficult to observe. That said, computational difficulties start in the measurement device, that's where you should do collapse if you don't want exponentially complex calculations.

Outcomes of observation are explained by linearity of evolution operator and decoherence, but then you get human factor: different people have different problems with this explanation, I think it's because unitary motion is unintuitive, Aristotle knew explanation of heliocentrism, but it didn't work, because geocentrism was more intuitive for him. Some apparently suggest that to explain human experience you need to calculate mind from physics, which needs to solve the hard problem of consciousness, which is impossible.