I just listened to the latest Game Studies Study Buddies on C. Thi Nguyen's Games: Agency as Art. I was only familiar with Nguyen's work through a 2019 article that I cited in my speedrunning literature review from last year. Even then I admittedly only read the paragraph that directly references speedrunning. During the process of writing that review I opted, in most cases, not to include mere mentions of speedrunning. However, I felt that Nguyen's mention rose to the level of an argument worth engaging with.
Nguyen writes: "speedrunning is an alternate mode of encounter with the material substrate of a game, and not an encounter with the work[…]Speedrunning the software of Super Mario World is simply a different game from the original”. What drew me to this quotation was how well it fit into the terms laid out by Stephanie Boluk and Patrick LeMieux in Metagaming, or at least my extrapolation out of Boluk and LeMieux. The "work" is the standard metagame (or what I call the presupposition space) while the material (or equipment) is the possibility space, which includes but nearly always exceeds the presupposition space.
Nguyen's writing on speedrunning is reproduced in his book with slight alteration. Cameron Kunzelman takes issue with Nguyen's claim because for him the "frame" of a videogame is its possibility space. The game is the game. However, if this were the case, if there were no standard metagame, then nobody would ever call a videogame broken or deficient for having glitches.
However, hearing the other side of Nguyen's argument I find myself no longer aligned with him, as he essentially appeals to the importance of maintaining the standard metagame. Nguyen argues that goofery in videogames, what I would call improvised emergent gameplay or what Bernard Suits would call trifling, is also not playing the game. As a child I can remember playing Super Mario 64 without any particular goal, wandering aimlessly around the hub world and other levels. More recently I've taken up 16 star speedruns of SM64, in which I haven beaten the game in under 30 minutes. Despite having experienced SM64 both "casually" and as a speedgame years apart, Nguyen would say I have never played the game at all, despite the fact that the goals of the game were practically designed to be ignored.
This ties into another friction I have with Nguyen. He writes that "When you speedrun Super Mario Brothers, the point is no longer to get the most points possible." (128) Later on he expands on this: "[the] scoring mechanism [in SMB1...]creates a quantitative measure of various game achievements along a single value scale. It offers an accessible, easily applicable, and quite precise method of ranking different performances in the game. This lets us create high score lists, compare our relative achievements without ambiguity, and declare with some finality who the 'greatest player of all time' is. Of course, we could generate alternate scoring methods or goals, as with speedrunning[...]But most players seem happy to evaluate their performances entirely with that built-in, off-the-rack value system, and accept the rankings it delivers." (197)
While I cannot speak for everyone who has ever played SMB1, I feel confident in saying that for many players the score is utterly irrelevant, or at least only relevant insofar as it encourages specific interactions within the game (collecting coins, defeating enemies, etc.). If anything one could argue that the game actually values the speedrunner's gameplay, since it gives additional points for each second remaining on the in-game timer. But in general I think there is a reason that SMB1's high score runs are less popular than its speedruns, and that is that the standard metagame simply isn't as compelling for high-level players.
Lastly, I think that Nguyen's statement that "You shouldn’t judge or review Super Mario World just by speedrunning it" is fair enough (128). However, I think of games like the recent Only Up, whose viral streaming success is specifically due to its affordances as a speedgame. What Nguyen unknowingly forecloses is the possibility for the standard metagame to be a speedgame.