Similar to how the subprocess symbol works in the BPMN models, it would be great to be able to configure symbols to be expandable and collapsible and if this was available in multiple model types. This would allow the viewer to expand to see additional nested/embedded objects, but then collapse to see a simplified view.
Another disadvantage of the current approach to expandable subprocesses in BPMN is that the view displayed is just a model preview and the objects displayed are not selectable.
Do you know Show/Hide assignments in Architect?
Or for nested objects Show connections of overlapping objects