Hi iamdar...
TiddlyWiki always had bidirectional linking. If you create a link in tiddler A to tiddler B, you could go to tiddler B, open "info" from the tiddler more menu, and see tiddler A under backlinks. So obviously, backlinking was pretty hidden and not very practical. Too many steps to get to it.
Roam Research made bidirectional linking more visible. All the references are at the bottom of each page on Roam. So, since Stroll was an experiment to see how much of Roam that I could replicate in TiddlyWiki, I also made b-linking visible.
I use backlinks mostly for quick navigation to related concepts, but it also helps the serendipitous discovery of related concepts.
In the example you give, you wouldn't technically need backlinking, since each tiddler refers to the other (Country A and City A, for example). But as you hint at, there are plenty of topics where you might not want to go through all the effort of writing out each description fully and with ample backlinks, as you have done. Life is short, and hopefully you are not trying to compete with Wikipedia for completeness.
I tend to create what I call index tiddlers when I want to create a table of contents on a given topic (say, Index: Country A). I keep that separate from Country A, where I keep the definition and description. But I usually find a way to hide index tiddlers when viewing in context or as transclusion, because index tiddlers tend to be longer.
If you haven't seen Drift, you might also want to see the way it handles references.
https://akhater.github.io/drift/ It was inspired by Stroll, but avoids the extra column, and adds tagging and freelinks to the references, which is more thorough.
I don't think the focus on backlinking in Stroll is pulling TiddlyWiki away from its roots. In part it is just making more explicit the backlink feature that already existed in TW. And in part, because ultimately TW is about customization and mixing and matching to create the workflow you need. And that would be my main advice to you as you build the world builder: don't rush to make it public. Play with it for awhile, use it yourself, and tweak it in order to eliminate defects and make it work for your workflow. That way when you do feel comfortable with the 'final' product, you will be able to describe not only the features but a suggested workflow. You need to be able to articulate what WorldBuilder does best, and for what real world use cases.
Not offended, I understood what you meant, and I hope the above helps.