This patch caught what, according to the cited RFC and the IANA Media Types registry, is a typo. This is an easy one to forget.
Should the API in FacetFile raise a warning (warnings.warn?) whenever image/jpg is used?
A solution in the ontology for this ambiguity issue was proposed in UCO Issue 363, but it is stalled on some logistical and design matters. Programming language-level review is another option for preventing this ambiguity, though it would be an encoding beyond what is in the ontology.