Libdoc with new style for arguments - #5734

I like a lot of what is here. Think it covers René's concern about unused space. I also think it puts this in a good spot for a release now and give people a chance to try it out and comment about it.

I personally like the current styles a bit more that this. The current table has some wasted space, but I consider is easy to understand compared to this version.

I find it a bit strange that the data is kind of formatted as in a table but kind of as in a list. I would think using either a normal table, like we now have, or a list, like this mkdocstring example does, would be simpler. That said, I see problems with the table approach and I also see that complex type hints and long default values would probably cause problems with the list approach.

The above said, I'm not really against these styles, especially when the comments here have been positive. I'd like to hear opinions from a bit wider audience, though. Could you @aaltat and @Snooz82 create examples with the current styles and with these PR styles so that people could test them? It would be best if both versions would be somewhere available so that people could experiment with the real docs instead of just seeing images. Using a real library like SchemathesisLibary or/and BuiltIn would be best. If majority of the people like the new styles, and we got at least few votes, I'm fine with using these styles in RF 7.5. We can, and probably should, do a bigger Libdoc HTML update in the future anyway.

If we go with this design, I'd somewhat strongly want to remove the equal sign from the default value. I didn't like it too much with the previous design either, but back then the default value was after the argument name and having name = default <type> was somewhat close to how this information is shown in the source. Now that the default is after the type, which I consider a good change in itself, the equal sign is just confusing.

Closing, repleaced by #5741 #5740 and #5739