The system has a few different objects that can be embedded in pages and checklists - todo lists, data lists, docs, files, comments and other checklists. The feature needs some work before its usable.
Understood.
I think the underlying problem here is that you have a particular model in your mind of objects and actions, and I, as yet, have no way of discovering what that model is, why it's useful, or how I use it. I think you need to make explicit, either through the interface itself, or with an explanation (big, bold cartoons can be effective) what your underlying model really is.
Whether you should stop coding and do it now, or whether you should finish alpha coding and then do it, is up to you, but without some sort of communication of your model of what happens, I don't think you'll get any useful feedback.
I hope that helps - it's intended to. More via email.