Yes, I agree. And the subjectivity of the macroscopic domain is neglectable when it comes to small subsystems.Detectors are usually all there is relevant in the sense of measuring a particle. They are indeed big enough to be treated classically. There is nothing more subjective in the quantum description of Nature than in the classical description.
But problems appear when the "quantum part" contains parts that are treated both as quantum, and as part of the classical part (ie. part of the observer). And even worse when one considers gravity.
The problem I see is this: Unless we find a more coherent theory that connects or relates the effective theories (much better than today that is), then - in principle - the scientific findings and the inference for any given experimental setup, has to be done independently for each observation or energy scale. This is not "wrong" but it leaves us with a patchwork of effective theories, with a bunch of experimentally tuned parameters, without the slightest clue how they are related.The question, which symmetries (or regularities, describable by mathematics) are manifested in Nature has to be found out by observation and model building in close interdepency. It's at the deepest level what the entire endeavor of the natural sciences is about.
Can we improve this situation? How?