I suggest only numbering expressions which will be referred to later in the paper. If "for future discussion" means a later paper, I think referencing equations in earlier papers solely via a number is a pretty good way to seriously discourage readership.