I am a little confused by this policy. These are not quadrinomials, they are just typical subspecific trinomials with added (and removable) classification connecting terms, per Art. 24 Ex. 1 of the ICN (https://www.iapt-taxon.org/nomen/pages/main/art_24.html). So, the names for the above examples are just Haloragis exalata var. exalata and Haloragis exalata var. laevis, not the form above with the classification terms. If that is followed, then it seems to me that the site shouldn’t need to implement any additional functionality to take four-term names, it just needs to allow forms to be grafted to varieties and varieties to be grafted to subspecies. These infraspecific names work the same way as species placed in infrageneric classifications; we don’t say “Lactifluus subg. Pseudogymnocarpi sect. Xerampelini persicinus,” we just say Lactifluus persicinus. If it isn’t an issue there, it seems to me like it shouldn’t be an issue here either.