In another topic I strove to define free agency as a category of event causation which is neither deterministic nor random nor a combination of those two. I would personally define free will as possession and direction of free agency.
You seem to imply, again, that both a clear definition and evidence are required.
This is either equivocation on the word “empirical”, or a false choice in the first instance. If an empirical claim is one based on empirical evidence, there are other categories of claims besides evidence-based claims and wholly abstract claims. Russell’s teapot comes to mind.
If an empirical claim is any claim concerning something that exists, then even my conception of free will above - which admits no evidence of existence, only evidence of nonexistence by predicting human behavior - qualifies as an empirical claim. And for Platonists such as myself, every abstract concept “exists”, and thus wholly abstract and empirical are not mutually exclusive categories.
~Max