To, I think, answer my own question, after watching the video here:
The “Search Engine for Projects” refers to a more controlled situation where a YaCy instance is used exclusively for a particular project where a project might have a wiki page, a forum, a blog, a content management system each with its own separate search engine. YaCy could be used to simplify that kind of situation.
What I had in mind was a means of doing that for all open collaborative projects across the entire internet, which would involve all related project pages using the same “project key” in their metadata string, which would allow any search engine to associate the pages.