Hypothetical versus Verifiable features

OSM focuses on Verifiability as a standard for determining what information should be contained within it. OHM focuses on historical elements that may be poorly documented or whose location may be fuzzy.

To this end, how should features that have locations that are known to be approximate be tagged so that their boundaries are not used as gospel by renderers. Similarly how should evidence / sources be linked to these instances?

Editing_Standards_and_Conventions currently suggests using source=approximation which somehow does not seem sufficient. What about using OHM:Hyphothesis?

Or if multiple concurrent possibilities exists, would something similar to OHM:Alternative record that a mutually exclusive feature possibility exists?