Jeff If people feel it is helpful to have a prefix for a heading (I find it to be cluttering when you have multiple levels of headings), then you could keep the current project prefix as the heading prefix and change the project prefix to something more indicative of a project.
I think many, including me, use Legend to manage tasks, among other things, and being able to group and filter by Projects, which may include not only tasks but emails, notes and other resources related to getting projects done, is very important in order to improve focus and efficiency.
Jerud Having to separate projects into a separate document as an alternate way to achieve the results of grouping and filtering prevents me from then also being able to see related non-project info together with project info, and limits sharing flexibility, even with the new proposed node sharing. In this case, being opinionated about item types, with related specific funtions, adds very important purpose-specific value.
You can accomplish a lot with a carefully planned system of tags and manual formatting (e.g., you could tag an item as #task instead of having a task type, or #project instead of a project type, or #high instead of setting high priority, or #star instead of applying a star, and then allow grouping by specific tags), but it would be very messy visually and difficult to maintain than purpose-specific item types with prefixes and would really discourage most users who don’t want to invest so much time building their own system. No, I think you need to have purpose-specific item types to fit the most common needs.
If you really want to not be opinionated and not be a purpose-specific tool, but rather allow/require people to build their own, then allow users to define types for which you can specify type name, prefix icon, prefix character, text filter character, groupability, etc. and allow them to assign from a set of specialized functions (Jerud’s superpowers). Then have presets that include Project, Task, Bullet, etc. While this might be good long term functionality, there are a lot more simple short-term improvements and bug fixes that should be addressed first, else this becomes another feature like mirrors… in theory, really cool and useful but in reality so difficult to get to work right that they are not, and trying to do so takes development time away from much more pressing needs.