iNat has a single global taxonomy that we all share so the system won't allow you to use Jepson's taxonomy and other people to use other regional flora's. iNat's reference for vascular plants is Kew's Plants of the World Online (which is the only global plant checklist at the moment). We can deviate to some extent to accomodate regional flora's like Jepson but keep in mind that these deviations affect everyone so while they work ok for endemic or localized species they don't make a lot of sense for Mediterranean species with global non-native distributions like this one where the same argument in reverse could be argued from another part of the community using another regional flora.

iNat has synonyms in that if you search for Festuca perennis then Lolium perenne comes up. But if you mean programing something where I could say to you 'thats Lolium' and somehow iNat could understand that by that I mean Lolium sensu Kew and that your understanding of Lolium sensu Kew actually might mean Festuca sensu Jepson and knows how and when to translate that so that to you'll sometimes see it appear as 'thats Festuca' and then how to maintain all of those observations on different taxonomies concurrently for each user's viewpoint and also train separate computer vision models for each user's viewpoint... then yes that would be a huge task ;)