It does look like the identifier here intended to park this at genus level (in which case, why not a disagreeing ID?) - the best way to confirm this is to ask them, via an @ or private message.
iNat prioritizes the finer ID (observation ID) for display, but it should be the community ID (still at genus in this case) which is sent to GBIF. It was the case in the past that the observation ID got sent to GBIF, not the community ID, but hopefully that error has been fixed.