The recent question "Is it OK to invent as I write, or should I plan the entire story first?" seems very closely related to "How much planning should go on before beginning writing?". Since the answer to the more recent one and the answers to the older one do not overlap, I do not think these are duplicates. (The specific questions themselves are clearly not duplicates. The older is asking if an outline is enough; the more recent is asking if writing can begin with almost no preparation.)
However, I feel there is enough overlap in what a good answer would provide (showing the range of preparation various writers make) that a single question could address this in the broad context that is implied in the title/question summary of the older question.
This broad question could be supplemented by more detailed questions along the lines of "What are the problems and benefits of writing with minimal preparation (pantsing)?", "How does character-driven plotting work?" (with text indicating that what is sought is not just a mechanical description but an outline of the tradeoffs; a similar question might be made for setting- and theme-driven plotting), "How are plot outlines helpful (and problematic)?", "When is world-buidling useful?" (a similar question might ask about preparing character descriptions), "What challenges are involved in writing toward a pre-established ending?", etc.
I suspect I am trying to force too much organization onto a questions and answers site, making it more like an encyclopedic reference (with cross referencing to handle various levels of detail and other relationships). Stack Exchange's competing answers method also seems to make broad questions more problematic in that individual answers will often only address one aspect and lean toward "bad subjective" (as exemplified in the older question) and people seem hesitant to incorporate others' content into their answers (even with Creative Commons Attribution Share Alike licensing the scoring aspect makes such seem a bit like stealing).
Side thought: I wonder if overview would be useful for distinguishing broad questions seeking general, high-level information (making "here's one solution" answers into non-answers since they do not actually answer the question). Sadly, such makes getting multiple answers less likely since requiring a long answer tends to reduce the distinctiveness of answers (a little content might be different and the organization, perspective, and style could certainly be different), which reduces the incentive to answer beyond just the effort required for an extensive answer (as the first decent answer will provide most of the benefit of a good answer). This effect might be compensated by providing badges for posting decent (by score) long answers to overview questions. (One might even be able to change the tag wiki into overview questions. While that would increase the incentive to provide some of the information that would normally go into such a wiki, some of the tag wiki information is more Meta-oriented, indicating what types of questions are appropriate and how to express them.)