Modifying best practices for older, complicated, somewhat messy DAGs

I was hoping to get some advice on organizing a complicated, sometimes-messy DAG.

I’m working on a dbt project that’s been around for a few years now and has had many people working on it. It’s much more complex and messy than having a a structure where every data stream goes through staging → intermediate → fct/dim structure.

Has anyone implemented a more robust dbt organizational structure that uses additional model categories, or otherwise simplified a pre-existing dbt project?