Yeah I mean as a statistician I don't believe we live in a world of "facts" and instead in a world of
evidence and
probability. Form my perspective having a reasonable (but not complete) level of confidence, or being able to reasonably falsify something (within a respectable margin of error) certainly constitutes as
knowledge even if it's not fact. I don't think of the physical sciences (often which are founded on experimental science) as a "repository of facts" but rather as a
working model of the world, which has high predictive accuracy. It's like that saying from
George E.P. Box "All models are wrong, some models are useful." Even the models for motion break down when you either go very large (motion of galaxies) or very small (motion of atoms) -- one of the bigger open problems in physics is unification of these models.
But even
math itself is not perfect and is really only a
language which is convenient for describing
observed phenomena. I'm not sure if you're familiar but for example
Godel's Incompleteness Theorems, which fundamentally rocked foundations of mathematics at the time, which shows that any axiomatic formal system can be either complete (in the sense that anything that is true in the system can be proved within the system)
or consistent (in the sense that you cannot derive contradictions using the system) but never both.
"First Incompleteness Theorem: Any consistent formal system F within which a certain amount of elementary arithmetic can be carried out is incomplete; i.e., there are statements of the language of F which can neither be proved nor disproved in F."
So even something which seems so dogmatically "factual" as pure mathematics is itself incomplete -- we know that our standard set of axioms are consistent but consequently not complete but even worse there's no
better system that doesn't also suffer from this problem!
So I tend more to the perspective that physical science, and indeed mathematics (and consequently statistics), is not fundamentally a "collection of facts" (nor is knowledge, imo, limited to being a collection of facts) but rather they express a working understanding (in forms of systems, models, and results) which have either
yet failed to be falsified (e.g. gravity) or which are particularly effective at
predicting observed phenomena (e.g. models for motion) or occasionally both.