A "workflow" involving many steps may make eminent sense when you talk about a corporation or any other large organization produces whatever it produces. The question for me is how much sense a, say, four-, five, or seven-step fixed workflow makes when you talk about efforts individual—especially as far as writing and note-taking is concerned. I am beginning to think that it does not make much sense. Yet, I read more and more about such supposed workflows, each involving a different application. I will not refer to any particular discussion, but here is one example.
- Taking first note and ideas (nvALT, and possibly paper)
- Outlining these ideas in some outliner application (opml would be a good option)
- Transferring the outline to some mind map program to "flesh out" these ideas
- (Re)converting the mind map to a text document to work on it in a more capable editor (or perhaps even a more capable outliner)
- "Writing it up" in an rtf or html editor and final storage.
Therefore, the fewer steps there are, the better designed the approach is. At most two, I would say, and perhaps three. But perhaps what I am really saying is that the use of two applications is better than that of many. I can see, for instance, how someone may brainstorm in Scapple. Import it into Scrivener, fool around with these ideas, using the outline, the editor (and perhaps even the fake index cards), and then compile the final product. That would be two (or at most three) steps. The important thing is that you can constantly switch between "writing up," outlining and visualizing the materials without having to switch applications. It's just that there is no clearly defined sequence of steps that must be followed in a linear way. And I consider that a good thing.
I have used the example of Scrivener as such an integrated writing environment, but there are other applications that allow you to do this. As I have said many times before, I use ConnectedText for everything and convert to rtf at the very end.