A four-level hierarchy like this one, where the real variation comes at Level 2, with a railroaded top-level plot and two lower levels of fundamentally cosmetic choices, is certainly going to be a commercially efficient way to produce interactive fiction.
And sometimes IF needs to be efficient, like when it comes with graphics and voice acting. I really enjoyed Telltale’s first Walking Dead game, which basically follows this pattern. I also enjoyed the Heroes Rise trilogy, which doesn’t have a whole lot of “real” branching but a whole lot of cosmetic “how do you feel about that?” fake choices. There’s a huge audience of people who read their CoGs just one time, who come to IF more for the sense of freedom it gives than to actually explore the variety on offer. Efficiency works great for that audience (and that audience probably buys more games overall than customers like me who read most games 3-5 times, and our favorites rather more, so it’s good business for CoG to aim there too).
At the same time, it seems to me that with ChoiceScript we trade the artistic/storytelling possibilities that are tied in with voice acting and graphics for – among other things, but probably first among them – a greater ability to produce real variation in our interactivity.
We can write stories you explore, much much less railroady than the map described by Episode, stories with odd corners to discover and factors that play into each other in emergent, hard-to-map ways throughout the story.
Choice of Rebels definitely has a Level 1 backbone plot. Your rebellion starts with you disrupting a public execution and fleeing to the greenwood; you bring your band through a hard winter, meet a few people in the spring, have an army come for you in the summer, and end up on your way across the border into the wilderness. None of that varies based on player choice. I wasn’t writing a rebellion sandbox. Them’s the rails.
But there’s no clear hierarchy of branches below that. There are lots of ways through the thicket of the story; there are characters you may or may not meet, and most chapters have extensive pathways that many readers will never see. The handful of basic stats interact with chapter-specific challenges to produce overarching outcomes that you can’t easily map (some of which have, to my surprise, taken dedicated readers literal years to hunt down).
That’s less efficient to write, obviously, but I found the results very satisfying both as a writer and a reader. I don’t mind my IF on rails, but the stuff I love, like the COG work of Kevin Gold, Kreg Segall, and Kyle Marquis (the three K’s!.. no, wait, scratch that) roll out an elaborate world to explore. I think Kevin, Kreg, and Kyle have each found ways of doing it much more efficiently than I do – there’s much more method and less messiness in Kevin’s process, certainly. So if you’re going to code dive, look to Choice of Robots and Choice of Magics rather than Rebels. Compared to the GDC/Episode schematic structure, I think Gold’s games offer a much more satisfying model for writing IF.