Need input from the many projects that have aimed to improve upon wikipedia already. Citizendium, Handwiki, and Sanger / Knowledge Standards Foundation / Encyclosphere, even self-admittedly biased ones: Rationalwiki, Conservapedia.

I think the issue is not to aim for a variety of contributors, but to allow the crowd to sort it out, and to allow for user-side filtering. It is the same challenge as with algorithms - allow for a market of filtration of data. Do not aim to restrict data.

With an encyclopedia page - the basic model would have multiple versions of the page available to the reader, with the reader choosing whose version to trust based on presets, preferences, defaults (be careful!) and the advice from peers.

The more advanced model (which has some effort and discussion into it already in various places around the internet) would involve realtime dynamic population of generated text based on user choice and bold, italic, or differently colored fonts to show various levels of trust of the information provided, based on user's preferences, presets, or the preferences of trusted peers.

Now, there are possibilities to have the page generated in realtime by ChatGPT-like services and data vending machines...