I've had similar thoughts in the past with blogs. Make more controversial opinion pieces require readers to indicate their concern and feedback, and then address inline. This way, I can understand the percentage of readers that bounce after reading the headline or don't engage too deep.
I have in the back of my mind to do the same. I think there's a ton of potential there, because you can pack a lot of nuance into an experience where only the nuance the reader actually cares about is surfaced. Additionally, doing the research to support objections both enhances your own understanding of the situation and shows the reader that this is an avenue you've considered.
There's one UX wrinkle I haven't sorted out yet, which is that I would ideally like to let readers propose new branches to support rebuttals/perspectives I didn't consider in the initial draft. Probably the cleanest way (particularly in my case, because the source is open) is to request PRs on the Ink script with new branches. But non-technical readers are unlikely to be comfortable working with PRs, even in a language as approachable as Ink.
Well, and a second, which is that in my ideal UX for something like this, I'd like support for explorable models (like in Jupyter). You could probably merge Ink and Jupyter because Ink's Javascript integration is open (I added support for "choices" that behaved as hyperlinks in my portfolio site linked above).
To that end, these days I'm exploring hacking the Godot editor into a browser of "model-driven documents", but life is busy so progress is slow.
