That is an interesting question. Am I allowed to say "I don't know"? Because I don't know. And how could I know - and if it were sufficiently isolated, how could anything else "know" (aka interact with /respond to/experience/perceive/detect...its presence)? And if its presence could not be detected (etc.) so that in fact the entire (rest of the) universe were completely unaffected and unaltered by its presence, so that, in fact, this isolated atom could make no difference at all to anything at all, in what sense can it be said to "exist"? OTOH - the very fact that it is said to be "isolated" and "in deep space" suggests that it does indeed exist because its isolation is a relational aspect of its reality - and in any case, it is not really clear yet that space itself is not a "thing" - is it?
An even more mind-bending thought experiment is to imagine that this isolated atom (and by 'atom' I mean an indivisible bit of reality) is the ONLY thing that exists and the space in which it exists is exactly that - space - nothing (and definitely not "something" for the purposes of this game). Now, how would it "know" which was up or down, whether it was stationary or in motion, whether it was spinning and in what direction, whether it was now, or before or after...if there were really, truly ever only one thing in existence with absolutely nothing to relate to temporally or spatially (or any other how), in what sense could it be said to "exist"?
Regarding your 1st para, we do in fact have a situation fairly closely approximating this in deep space, where atoms, ions or molecules are separated by such distances that they interact with matter or radiation very seldom. (The evidence for this comes from the interactions with radiation that they do occasionally undergo and which we then detect.)
I think therefore I could argue the reason such an isolated atom can be said to exist is because it always has the
potential to interact with other material entities, whether or not it is doing so at any given moment.
Indeed, pursuing this line of thought, if one were to contend that existence requires interaction, then one would, rather absurdly, be implying that ordinary everyday matter is continually winking in and out of existence! After all, interactions occur only at intervals, and in between one has matter and radiation in states that are
not interacting.
I don't myself see any purpose being served by such a complicated concept of "existence". To me, the concept of "existence" ought to meet some sort of Ockham's Razor test of practical utility.
Regarding your 2nd para, I see your point but do not see where such a speculation gets us, given that this is not the case.
Regarding panpsychism, I can quite see that if the idea of "mind" is defined merely to be the process of interactions of matter and radiation, then that nicely gets rid of the problem. The human mind can be seen simply, and correctly in my view, as just the activity of the brain, i.e. the interactions of neurons and the chemistry therein. All other "mind" activities of other organisms, consciousness etc, can be treated likewise, as interactions of material entities. So we have, at a stroke, got rid of dualism and reduced "mind" to a physical basis. But then I do not think I would call that "panpsychism".