The user may simply want a place to track all of their academic process, and ideas/interests. (Like with MAL and Kitsu, no need to share a profile until does.)
Therefore, the minimal product should be able to self-host. If the user wishes to enter into the social dimension, they should be able to merge their data into the larger (central?) database.
If they wish to share the list through their own avenues, for exaple their own website, that should be fine.

The user is able to request suggestions through identifying a certain domain (certain group, users that share that [or similar] entry/ies, etc).
Likely implemented by pushing their request to clients of the "certain domain", and eventually as patterns form of what users get recommended, show that to the requester as well.

Collect similar resources (papers, books, videos, etc. of similar content or subject) into some kind of module or container, then create something about/with that module/container. For example, research on Ranma 1/2 may lead to history of gender discussions in whichever localities, or to the history of martial arts comedies, or something about the classic myths & folklore referenced and played, etc., all which depend on the student to determine best scope of sharing their understanding through a single lesson, or as a "course" (meaning many lessons over a length of time).
After the lesson is created, able to share with public, though designing and publishing a lesson outline, youtube lecture, blog post, interactive, etc. Ideally, able to form into a syllabus, with outline, goals, integration of materials, etc. and able to self-reflect by comparing to other syllabi.
As more lessons are planned (if topic seems to be shallow yet specific), able to integrate into a course (by choosing what lessons to add, and how to integrate and organise lessons). If decide to define a course as a goal of progressing through research, user will be able to self-guide or utilise reccomendations in order to fill course with content.

At this stage, basic structures of playing at an "online university" (without accredidation or hope to be taken as a "legitimate" education for anyone outside the project) are in place. Users represent themselves as students, and through developing on their own, by exploring curiosities in useful ways, become "experts" in their particular field - enough to socialise with other people that know and understand the same subject, so that everyone involved learns, and is able to share and push understanding further.
As courses are designed by individuals, able to collaborate to refine their own courses, and to ultimately consolidate understanfing of that into a single course. (For a particular target audience, of course there will be multiple levels and languages, etc.)
To do so, establish ways courses can join into "departments" that represent their field, aligned to typical university structure. (ex. "Physics" usually includes multiple research areas, but able to distinctly differentiate from the adjacent sciences. Also, emerging fields like "Veterans Studies" (specifically "veterans in society") typically are considered "Cultural Studies", but tend to differentiate between "cultures" - by geography, history, etc.)
More research required on self-adminstration, probably with establishment of trusted moderators. Probably the same mods that approve merging a user's list into the larger database. (Users able to constantly evaluate trusting mods, and mods able to understand the development of users.)
Implementation likely moves toward an LMS for teaching at distance: due to the natures of time, of life, and of the internet, probably asynchronous in most cases.

https://www.youtube.com/playlist?list=PL09go9ZyhT_7zwfV7DwuhvOiGEFLrutAD - ProCrastinators Podcast Lectures, also Digi After Dark hour-long rants. Very informative. https://web.archive.org/web/20110630035746/http://studentscircle.net/live - StudentsCircle and OCW https://arxiv.org/abs/chao-dyn/9909001