Taxon nodes, ranks and hierarchy dependence

Our collection has many agricultural holdings including hybrids and cultivars (typically noted with apostrophe ‘cultivar’ notations. The cultivars can be associated with many different taxonomic ranks, for example;
Malus ‘Adams’
Malus baccata var. mandshurica ‘Nikoensis’

Now, through my own ignorance, or perhaps the 10-year-old version of Specify that we first started using, I created 2 different levels of ‘cultivar’ ranks in our Taxonomy tree. One at the level ‘below’ species (Cultivar(sp)), and another at the level below subsp./var. (Cultivar(ssp/var).

Recently, I was reviewing a related hack that I implemented, where I created an artificial species-level node to affiliate cultivars which didn’t have any species ranks indicated:

I was trying to figure out how I could better capture this, and to my DELIGHT! I found that I could simply edit the taxon to still be a ‘cultivar’ and keep the ‘cultivar’ notations that are important, but that I could assign a genus as its parent, and not need the [sp.] placeholder at all! (See ‘Adams’ circled in blue in the image below) Amazing! just what I should have done the first time.

For example, Malus [sp.] ‘Albright’ (from the image above) can become Malus ‘Albright’ (which is much better).

So with that in mind, what make sense for Malus baccata var. mandshurica ‘Nikoensis’ (selected in red in the image below)

Is it possible for me to use the same Cultivar(sp) taxon rank for this, and to simply have its parent be the variety? Instead of having a second (Cultivar(ssp/var) rank? (how would this display? would it break the whole tree?)

In my opinion, this isn’t just a cultivar problem, but also exists in several infraspecific taxa, where some determinations ‘skip’ some ranks, and others don’t.

Eg. How do I best include a specific forma, if it has ‘several possible parents’. I have always just duplicated the child-taxon as required, but wonder if that is the way that makes the most sense?
eg. A determination could be Genus1_species1_forma123, but the same forma node could also be included in a Genus1_species1_variety1_forma123.

Is there an recommendation to an alternative to having forma123 in the tree multiple times with different parents???

Thanks for any insights, or comments on how others are handling these options!