Perhaps for some perspective: the scope of a "name" can be limited, that is, "local", or it can be "global". Even fairly basic programming principles nowadays (with virtually unlimited namespace, as opposed to 1960's Fortran) recommend keeping namespaces as local as possible...
... and this advice applies to most mathematical notation, as well. That is, there are really very few symbols/notations that are understood globally and unambiguously. That's fine. I think one should be reconciled to (re-) setting the "names" in a given "scope", and not at all depend on document-dependent notation/names as though they'd be something the reader would have assimilated. (Excepting perhaps a very small number...)
Even then, if "tau" has two different and conflicting senses in the literature, there is no genuine intellectual/scientific purpose served in being mute on the point, and trying to find a way to dodge the operational ambiguity. That is, in each of your sections or subsections, tell what "tau" is. Done.
That is, do try to find a viewpoint from which this is a non-issue.