I'm afraid Mr. Nelson has not made quite the point he thinks he has made. He is (rather flagrantly) abusing equivocation, hinging on the ambiguity between two senses of "bad" or "undesirable" design. This argument is at high risk of becoming a motte-and-bailey fallacy if not scrupulously addressed; I mention this only to cut off such responses at the pass, so to speak.
The first sense of "bad" or "undesirable" design is "design which is not strictly favorable to the player and which may create an environment of hardship or a perceived or real bias against the player." This is "bad" design only in the sense that it is design which intentionally creates hurdles to be overcome. But since the whole point of gaming is to do something (getting back to my four part typology of game-purposes!), intentionally creating good hurdles, hurdles that are interesting and compelling and fun etc., etc., is in fact exactly what I would call GOOD game design. Designing such hurdles so they are neither too high nor too low, neither too close together nor too far apart, etc. but rather sitting in the context specific Goldilocks zone, that's difficult, and we recognize when a game has achieved this. (For example, Hades received massive, and well-deserved, praise for making it so death is actually interesting and even rewarding, such that even someone like me who HATES most roguelikes and roguelites thought it was an absolute blast.)
The second sense of "bad" design is exactly what it says on the tin. Such things are often called "clunky" or "janky." They either don't work right, or require unintuitive and cumbersome actions for minimal benefit, or mess up through no fault of the player, or otherwise break down in some way. This kind of "bad" design is undesirable because it detracts from the gaming experience without adding anything in return. It absolutely should be avoided, and there really is no excusing it.
Mr. Nelson, possibly without realizing it, wants us to accept the former as "bad" design, so that we will embrace usage of the latter. This does not logically follow; if condensed into an actual logic argument it becomes a fallacy of four terms, which is an actual formal fallacy rather than the informal fallacy of equivocation. E.g.:
To fulfill the purpose of making a game is to give a good game experience.
"Bad" (read: intentionally difficult) design is sometimes necessary for making a good game experience.
Therefore, "bad" (read: clunky/janky/nonfunctional) design is sometimes necessary for fulfilling the purpose of making a game.
This is a formal fallacy because the argument as presented is flawed due to its structure alone, rather than its content. It is called the "fallacy of four terms" because the syllogism does not actually connect the 2nd term ("bad" design as intentional difficulty) with the 4th term ("bad" design as clunky/janky/nonfunctional), when normally those two things would in fact be the same term. If we use more specific phrasing the failure to connect the premise to the conclusion becomes clear:
To fulfill the purpose of making a game is to give a good game experience.
Intentionally difficult design may be necessary to give a good game experience.
Therefore, clunky, janky, or nonfunctional design may be necessary to deliver a good game experience.
Now, the fact that I have pointed out a fallacious argument does not, in itself, refute the key point being made. It might be the case that clunky, janky, nonfunctional design could be necessary to produce good game experiences for some other reason. However, in this case, I think it is rather clear that "intentionally difficult" is fine while "intentionally broken" is not fine. Broken things are, in general, almost always more difficult to work with. That does not mean they are somehow a better choice than non-broken things which achieve the same ends.
To give examples specific to Mr. Nelson's responses: having a trade market that can shift on a dime and sometimes "cheat" you out of a good deal is not clunky or janky design, it is in fact functioning, and in so doing it represents a real danger that high-risk trading folks encounter on the regular. Capturing that frustration for the player some of the time is, in fact, desirable. But if it's happening every other trade, that would not be desirable. It should occur often enough to be a real risk, but rarely enough that most people only see it a few times a session or the like.
Conversely, the example that the player is beholden to the requests of literally all clients, even if those clients are scammers or the like...no, that does not strike me as intentional difficulty. It strikes me as forcing players to do self-harming things solely because the game doesn't let you choose otherwise. That is clunky at best, and does not actually represent the fiction of being a trader in a high risk exchange market: no trade market has ever been set up where you HAVE to accept ALL requests no matter who or what they might be. That would be a massively abusable system and would essentially guarantee no one would willingly trade on that exchange.
Intentionally difficult design can be good when used judiciously, as shown by the outstanding success of Elden Ring. Intentionally broken design is not good, and should be avoided.
