Fibery already supports relationship attributes, right?
In my very recent experience with Fibery, most of its relevant components already support relationship attributees, particularly when compared to other Online Work OS platforms where they seem to forget that relationships can have attributes within the context of a relationship.
In Fibery:
- the formula columns can access any related entity along the relation chain.
- the formula return data types can be entities or their attributes such as dates, date ranges, numbers, locations, etc.
- the formula columns are usable in filters, grouping, sorting and reporting.
- and relationship filters work on any field or entity in the relation chain, to allow subsets of a relationship to be associable in different columns.
I am relatively new to Fibery, but as someone with a normalisation obsession 
usually hovering between 3NF* and 4NF*, who cannot resist extending relationships with attributes on junction tables, and enriching simple values (i.e. options in a select) with extended attributes on reference/domain tables, it seems like Fibery covers the need, except for parts of the corresponding UX.
Please tell me if I am missing something. 
* 
Sorry for the jargon, I have linked them to their Wikipedia article, just in case.
If Fibery does support this, then is UX the issue?
The issue for me is more a matter of UX, where instead of setting up separate junction tables, relationship filters, formulas, etc., we could just set up the relationship and its attributes in an improved set of UI components, seriously advancing the UX of setting up and using these attributes.
An example of UI changes to improve the UX: