I would add that it is indeed early days for Obsidian. This is both exciting and a pitfall.
The excitement lies in its tremendous potential. With a plugin community and its extensible nature, obsidian can evolve to have many diverse strengths. As plug-ins continue to develop, I suspect obsidian on one user’s machine will be quite different from obsidian for another user. Students and academics (even in the same field) work, learn and create in very different ways.
The pitfall is that most academics are trying to get things done today (well, usually yesterday). Academia might be centuries old, but neither Obsidian nor the user is going to be around to see how knowledge management evolves over that time scale. The practical question that software tools need to address is: what can it do for me today with respect to writing this thesis, grant proposal or manuscript? From that perspective, one debates (I certainly do) whether it is a good use of time to use Obsidian vs something that’s been around the block (e.g., org mode, Scrivener, DEVONthink). People have different tolerance for growing pains. No universal answer but something a user approaching Obsidian (or any other tool) should probably ask before jumping down the rabbit hole.
With regard to knowledge management vs writing, the two are so intimately connected, I (personal opinion) tend to favor a single app that does both well. Why dilly-dally with two environments when a single one can do both? One app is better than two is especially relevant with these classes of apps because they often require significant tailoring and personalized decorations and settings (plugins, keybindings, even some coding). The time invested in getting into one of these app ecosystems is not trivial, and the longer you can manage to stay in a single system the more reward you reap from your early investment of time and effort.