That is an interesting metaphysical objection. Does a particle exist, if you cannot measure it? There is certainly no reason to think it does. But is there reason to categorically deny is existence? Very hypothetically, you might build a "Laplacian supercomputer" that could answer this question. But until this can be done (wich I doubt very much), the agnostic approach, to leave the question as undecidable, is prudent here.
>> IIUC everything in the observable universe is sufficiently far away from anything else that any causal relationship with anything
I don't think that is correct. Suppose you have a particle that is very very near the horizon. This particle has its own observable universe and we are at the edge of its horizon. Now, this particle can have casual contact with the other parts of the universe (the parts beyond). But by the time the interaction takes place, it has already slipped out of our horizon. So from the perspective of the particle casual contact is possible. But from our perspective: Not even a chance to detect this interaction.
But what does it mean for a particle to be near the horizon? The horizon is ours, thus the particle must be near our horizon, which means either the horizon we might observe today, sans opaqueness of the initial universe this means a point in spacetime before particles even existed. Or the furthest point we might observe in the future, which means sending the observer at C in any direction. Now, would the observer ever reach a point where she’d be able to observe any trace of interaction with anything that wasn’t observable from earth at the start of the journey?
Let’s take a photon, or graviton, traveling towards the observer such that the observer would interact with at a distance from here expanding at C, that would be our particle “very very near the horizon”. It would never reach earth so strictly speaking it is outside our horizon so in a sense it is proof of existence outside the horizon. But let’s expand the definition of horizon to include all particles that could be measured by the observer, could any one of them have a history including an interaction with a particle originating outside this expanded horizon?
>> IIUC everything in the observable universe is sufficiently far away from anything else that any causal relationship with anything
I don't think that is correct. Suppose you have a particle that is very very near the horizon. This particle has its own observable universe and we are at the edge of its horizon. Now, this particle can have casual contact with the other parts of the universe (the parts beyond). But by the time the interaction takes place, it has already slipped out of our horizon. So from the perspective of the particle casual contact is possible. But from our perspective: Not even a chance to detect this interaction.