Yes. Prototypes are also useful. They can even start with no customisation.Why do that? Well it’s easy t o, for example, change the shape or colour of 20 notes if they all use the same prototypes. also as those attributes are inherited the changes are non-permanent. Thus you can use prototypes to start to tease the Y things from the Z things even if what Y and Z are isn’t clear beyond the fact they differ.
I’m also cautious about drawing lots of link lines too early. The issue isn’t that the link is not valid, but that 2 icons joined with a line have more (unintended?) visual impact than two not so linked. Note that you can toggle the visibility of links (at per-type scope) so that two can be a way to use links but reduce their initial impact.
Personal styles differ, but besides shape and colour don’t overlook subtler clues: badges, borders, flags (which can also encode simple attribute data).
Using short titles for notes (the full title can initially always go in $Text) helps keeps icon small allowing more space and less textual noise.
I also recommend Mark Bernstein’s ‘map grammar’ TBX shown at the 12 Nov 22 meet-up (the tread includes a link to the TBX). There are lots of simple placement styles (many of which don’t need perfect gridding to work) that can help surface relationships.
Theres’ no one right way though, as personal styles vary.