The concept you touch on in your quoted paragraph is something I've been fond of calling "cohesiveness". As of recently I've adopted the thinking that a single vision, and most generally a single person, should hold a position of authorship over a game. This isn't always a writer, as the game isn't always narrative in nature -- my friend Beau Blyth, lauded for his skill with multiplayer games, creates the code, art, sound, music and game design in his projects. Because one person with competence in all these areas is creating the game there is a resonance between all the separate elements. The authorship creates cohesiveness.
Beau and I worked together and have begun to move in different directions. While his interest is in creating multiplayer gameplay that can create social experiences between friends and strangers, I am fascinated by single-player games which feature exploration and narrative. Which brings me to the matters you discuss in the later part of your essay.
Firstly I should say that after your discussion of the former point I had some difficulty understanding precisely what your thesis was. Perhaps simply that more discussion of interactive narrative should be fostered. But you mention narrative creating a system of constraints to the player's agency in much the same way as the rules of a game do. This is a sensible point. However, even a computer "toy" (that is, a program which facilitates play) imposes a system of constraints by virtue of the limited set of interactions and logic which are available, as does a reader-guided storyline.
I really like player-guided play, to be sure. I intend, certainly, to make it a foundational element of my narrative game. But I also don't think that restrictions on agency are invariably bad. It's interesting when players create their own narratives in open-world spaces, but not all players will have the courage or interest to do this. Those that do -- myself included -- will not necessarily find our travels in an open-world game more compelling than one that simply tells a linear story. The latter is more dense and thought-out than what can improvise as we play. If for the purposes of a good story I'd rather play
To The Moon than invent the exploits of Corbidor the Dragon Puncher in
Skyrim, then it seems the former is achieving a lot more storytelling per man-hour put into it.
So here's my thesis: I value exploration in games, and I have a narrative project for which I will act as author -- game designer, writer and programmer. My solution to an interesting narrative which doesn't violate the player's agency is to turn that narrative into an explorative space. Exploration, being a form of play, is narratively freeform. It has a strong connection with
searching, which is a goal-oriented activity which adheres to the structure of a game or a linear story. Thus I take a form of gameplay which is fundamentally
play, and I lightly build a
game on it. If the player wishes to approach the system like a game, she may seek out the object of interest and proceed through the narrative. If approaching it like a toy, she will explore every corner, finding along the way the necessary means of advancement and opening the way to new vistas. In either case an authored narrative is dynamically explored.
Tangent warning:
I'm used to exploration through a physical metaphor -- walking, climbing, swimming, flying, et cetera. I relish the oft-unnoticed
narratives of the physical world. But storytelling does not adhere to physical laws, and time and space present difficulties to the explorative conveyance of a narrative by means of of physical motion. What I will experiment with is an abstract metaphor that allows memories, stories and daydreams to be freely traversed. This is just one of several viable solutions to the aforesaid problem, though.
tl;dr Making emergent open-world games is not effective storytelling; exploration works really well as a vehicle for narrative.