Looking back over what you've written, I think we probably do agree more
than disagree. The main difference is that I distinguish between the
information from which a probability distribution is constructed -- which
may include physical properties of a system, to the degree they are known --
and the actual probability distribution itself.