I don't really have the know-how, either, but the OP for the original lyrics reloaded states that it's not an update of Steven's plugin but a total rewrite, and since he did use some sources from the original plugin and didn't use others, I'd assume the reason he decided not to use the sources that he chose not to use was that those sites are not compatible with the structure of the plugin.
In fact, looking at that OP now, he says certain sources are "not possible yet" in that post, and those include the ones from the original plugin that he didn't use.
If something has changed and you're able to make a working YML file for any of those sites, I'll be happy to add it into the plugin. I've been pretty clear that the kind of RegEx skill needed to create these files is not in my wheelhouse and life is kind of coming at me too fast in this last year to get anywhere close to learning it.
The great thing about Lyrics Reloaded is that anyone can add sites to it on their own and then offer the YML for others to use, too.
Wish I could be more helpful, but I'm really just an overseer of the plugin. My skills are such that I can do cleanup and maintenance but not really creation.
The two don't interfere with each other at all as far as I can tell. Why not just use both?
The architecture of the plugin allows to implement entirely different provider concepts side-by-side (in the codebase they are called LyricsLoader). The current concept (StaticLoader) relies entirely on the websites having a stable and predictable URL per song, that can be derived just from the limited song metadata that is available (which often can be done with surprising accuracy as you know).
I originally planned to implement a second loader implementation that tries to use search functions of the websites. so the flow would be like:
1. Construct a search query based on the song metadata
2. Call the site's search functions
3. Extract song URLs from the search results
4. Try to find a single best match in the results, possibly applying some fuzziness
5. Extract the lyrics from the matched URL's site as usual
So provider Yaml files would require a couple more options (search URL, search query filters, URL extraction regex) for these.
In case someone is interested in implementing this, I can assist, but I will not implement this myself as long as there is no native MusicBee for Linux.