I'm thinking of the game-as-artifact, which will can include mechanics, illustrations and examples, setting, relatings of play and so forth. Everything a player receives with the artifact and that can go on to inform their use (including their determination that the game can be used in a way that can satisfy their interests.)
While the various aspects of a "game-as-artifact" are relevant, taking such an
absolutely holistic view makes it a hell of a lot harder to actually talk about design in any sense whatsoever. Because now "game" means so many completely disparate things that all have to be integrating together at the same time, and all of them need to be interconnecting and supporting each other, and so many of them work by completely different rules.
Font and font setting and layout. Illustration and art and color design. Cartography and geography and climatology and geology and ecology. Politics and economics and sociology and psychology and religion and philosophy.
And not one single one of those things has anything to do with
game design, as in, the actual process of crafting
a game. It has everything to do with crafting a
world, certainly, but the world is, if you will, "post-design" in terms of building a game that has rules.
It's like asking someone to produce the full tarot deck, with all of the major arcana, all the symbolism and meanings and art style and (etc., etc.), all on top of the complicated rule interactions between those various cards,
from whole cloth, immediately after saying "I think I feel like making a card game." Yes, all those many different things are good for the game's impact on players, for its overall success and longevity, for its ability to inspire, and more. But such an overwhelmingly holistic view makes it a hell of a lot harder to actually examine any of those pieces. Yes, the holistic experience is what the player will have.
But we cannot design games all at once, from the ground up, to contain everything the player will experience. We must, necessarily, design them piecemeal, furtively, sectionally. It is essential that we finish the most utterly critical tasks first, and then the next-most-critical tasks, etc. until the whole is complete. And, unfortunately, many of the things the player will experience
first are things that genuinely need to be done
last if the game is going to end up actually good and worthwhile as an experience. Trying to
start your game design by paying artists to make really awesome art, and then
hoping that your design happens to pay off on whatever the artist felt like drawing? Not a good idea.