He is saying not everyone needs the system to help in the same way and that the most essential thing for sandbox to function as he is describing it is that core play loop. It isn’t like physical technology. The X card might help some groups but it might hinder others. It isn’t always going to be helpful for groups pushing the envelope (and some people find it can cause more problems). The same with any mechanical development. Some of the examples I have seen in this thread for good sandbox play, scale back combat for example, making it less granular. In those campaigns having more rules to help positioning during a fight would probably not help, but in a game where combat is more standard, positioning rules could help (but such rules could interfere with the core play loop for some people if they want a more flexible and open approach to positioning), same can apply to rules for overland travel. One man’s helpful mechanic or procedure is another man’s obstruction
Some of the examples of "better" also seem to really push a metacurrency, points that the players can spend or gain via their declarations. So if a player does X then they know what to expect because the response is limited to Y, possibly with some randomization. This is obviously true as well in D&D for combat, if I attack the goblin with my sword I expect that if I roll high enough to match or exceed their AC that the orc will take Y amount of damage. If that accumulated damage exceeds their HP threshold, the goblin is out of the fight. With of course explicit rules messing with expectations here and there.
But I don't want that kind of systematic approach to social or exploration encounters so games that have that kind of play are meaningless to me. I need it for combat because I would have no clue how to run it. Meanwhile I like that D&D takes a more hands-off approach to non-combat with GM authority defining the world and player authority limited to their characters. It's more immersive and believable to me to follow the D&D paradigm, which is something I value over that systematic approach,
Along the same lines I don't see collaborative world building to be any more of a sandbox than the GM giving me options so I can make an at least somewhat-informed decision about where to go. If I go for a walk in the forest I may have an idea of what's over the next hill, I don't help create the waterfall that I find. That doesn't affect my agency or my ability to make decisions about where to go.
To use the so-so analogy I don't see much value to comparing a drill to a hammer. They serve somewhat similar roles but they are different tools with different approaches, objectives and experience.