Thanks for the elaboration. It seems to me that experiences are:
Hard-to-eff, as a good-enough theory of what physical structures have which experiences has not yet been discovered, and would take philosophical work to discover.
Hard to reduce to physics, for the same reason.
In practice private due to mind-reading technology not having been developed, and due to bandwidth and memory limitations in human communication. (It's also hard to imagine what sort of technology would allow replicating the experience of being a mouse)
Pretty directly apprehensible (what else would be? If nothing is, what do we build theories out of?)
It seems natural to conclude from this that:
Physical things exist.
Experiences exist.
Experiences probably supervene on physical things, but the supervenience relation is not yet determined, and determining it requires philosophical work.
Given that we don't know the supervenience relation yet, we need to at least provisionally have experiences in our ontology distinct from physical entities. (It is, after all, impossible to do physics without making observations and reporting them to others)