I'm not sure I follow what you're saying about the inefficiency. I think the point is meant to be something like "getting your opponent in checkmate via actually moving your pieces according to the rules of chess is 'less efficient' than simply putting the pieces directly into a winning arrangement" or "getting to the finish line of the race by following the course is 'less efficient' than taking a shortcut". That seems like it's similar to what you're saying, so I'm not sure if I'm misreading you.

what would it take to make knitting a game?

Well, the trivial approach would be something like a race, so you're pitting your skill/technique/speed against someone else. I think the more interesting challenge would be how to make knitting part of the "core loop" of making meaningful contributions to play (analogous to the role that drawing plays in the drawing games I've been posting about). Some element of personal contribution probably matters -- when I think of knitting I think of the process of actually moving the needles on the yarn rather than creating or improvising patterns, if that process is "machinelike" in that it's hard to connect to a personal contribution then it may be tough to make it part of a "game move".

I think your example of "knit, then show off what you knitted" is maybe getting close to being a minimal game but it feels to me like there's no mechanism that makes sure that your "move" matters to the game state -- if people can just ooh and ah and then promptly forget about what you did it might be indistinguishable from you not doing anything in the first place. If there's some sort of handoff and "build on top of what you got" it looks more gamelike to me, a la Exquisite Corpse. I guess what I'd be looking for is an answer to the question "how does what you knit impact what happens in the next cycle"? If there's that's an answerable question then it's looking a lot more like a game to me.

Your example of knitting multiple patterns also feels like it's getting close to a minimal game to me, but (assuming I understand what you're saying correctly) I also think it's not quite there. My feeling is that for something to be a game you need some sort of human decision-making or ability-testing as part of "the machine" that is the game, where your example seems like an open loop where the deciding just sort of happens at the end.