@juliewolf this makes sense to me to let project leads and partners to create their own discussions as they see fit.
I think categories have better “ quality control“, meaning that you repeatedly select from a set and the spelling is always right, so that categories are more reliable for generating automated sets of blog articles. So I would use categories to describe a few critical, well managed divisions, and I would use tags more loosely. I suspect categories would describe who, what and where if relevant… the other way, would be to leave them uncategorized for now, and wait until someone wants to create a subset, and then they can categorize the one that fit the set.
Maybe it’s worthwhile to brainstorm a list of all the reasons why someone would create a blog post!? To describe a project, to provide an update, to share a piece of writing, to teach about protocol or etiquette. The posts could be reserved for more permanent chunks with higher information quality. For more transient information I would rely on discussion threads.
I think there is a risk of having an over proliferation of low quality information in posts. But that might be a personal preference, because I personally have a very narrow bandwidth, many projects and have an overwhelming, constant flow of information. Thus information quality is more important to me than the risk that someone is excluded or transparency is not maintained by not having all the information on everything everywhere all the time. Submersing people in an overwhelm of information is a different kind of exclusion. Overload is a different kind of obfuscation.