All your examples are ways to counter the problem of the lack of informations. We need the UI because we need to compensate the limit of our representation. Instead I envision the removal of the HUD when the representation will be so good and self sufficient that an UI will just be superfluous since the game experience already delivers in the first person all the feedback the player needs.
For example if we take a primitive 3D game with placeholder graphic, the /wave emote will be just a text message you can read. But it's when the /wave corresponds with an actual action of the character that the message in the chat starts to become superfluous.
The whole text chat is already an interface that tries to compensate the limit of the game. So, while the characters in that precise situation are supposed to communicate with their voice, in our representation we simulate it through text. Teamspeak here replaces the rudimental UI with a direct element. With a more faithful simulation.
The removal of the HUD isn't to hide informations to the players. It isn't to obscure the mechanics to encourage the roleplay as it happened in some games. It's instead the result of delivering the same informations in their natural form.
The ideal game here is the game that doesn't need any specific knowledge. Today the games are complex and aren't that easy to play if you don't have already a background as a gamer. They have their own culture and you need to be dedicate enough to it in order to get involved and have fun. So games in general need study and practice. These are HUGE accessibility problems. Firstly because only a small number of people share those competencies and secondly because the experiences they deliver are too detached and artificial to be enjoyed to their real potential.
The television and cinema are way more direct. In fact they are able to reach a broader public. Yes, you still need the specific competence to understand the cinematographic language but is simpler (and way more common) because what you see on the screen is "analogic" (similar) to your real life experience. So you can relate to a politics discussion, a war, a love story, a fight and so on. You have the competence to understand these stories because you share a common background on this reality.
Games don't offer completely alienated experiences, I wrote about this as well. They still use common experiences like fear, awe, conflict, love, surprise, mystery and so on. All these feelings are again part of a common background. And we definitely do not need to learn "interfaces" to understand them. Everyone in this world, no matter of age and culture, can dream one night to be in a forest and facing a huge orc. And there isn't anything special to learn to relate yourself to something like this.
But if this is a "game", only a very tiny group of people will have the competence to understand the representation filtered through this medium. Because you need to relate to a symbolic representation with its own rules and quirks that only vaguely resembles to the actual situation we want to simulate.
All this is superfluous. It's an useless weight. These are parts that will progressively vanish to favor a more direct experience that doesn't require the player to have already developed specific competencies about "computer games". And it's not just something I say as a personal opinion. But a fact.
If we erase this useless geekdom everyone will be able to relate and enjoy the experiences coming through the games. Hence a broader audience and a better accessibility for the medium as a whole.
This is a very common point of view for the technology in general. The goal has always been about making these tools easier to use so that you don't need a specific knowledge to use them. The technology, in fact, has a barrier to its use represented by the interface. We are moving slowly to remove these interfaces. Till we'll reach a point where the people will be able to "talk" to a machine to have it follow those instructions precisely. Without needing anymore to read manuals, learn icons, buttons etc...
Why? Because the machine itself is finally able to relate to the verbal communication that is what we naturally use (and have competence to).
Take the example of the /wave I made above. If we call a child to watch what is happening on screen when the emote is just a text message, the child will understand what is going on only if he already developed various, external competences. For example he needs to understand what a chat box is, how it behaves. He needs to understand that the the lines he can read are the lines spoken or performed by the characters on screen. All these informations aren't trivial at all. They are for all of us who played these games for so long, but not for him. They aren't so trivial for a broader audience.
Instead. If the model on screen performs the /wave action with a realistic animation, the child won't need any knowledge to understand what is going on. He will smile and tell you directly, "Hey, he's saluting you, look!".
It's not a case that so many people like the emotes and /dance and jump around. These are all "analogic" representations that feel direct. Without the need to parse the game and understand its obscure language.
Removing barriers
Oh, but that's not the point.
All your examples are ways to counter the problem of the lack of informations. We need the UI because we need to compensate the limit of our representation. Instead I envision the removal of the HUD when the representation will be so good and self sufficient that an UI will just be superfluous since the game experience already delivers in the first person all the feedback the player needs.
For example if we take a primitive 3D game with placeholder graphic, the /wave emote will be just a text message you can read. But it's when the /wave corresponds with an actual action of the character that the message in the chat starts to become superfluous.
The whole text chat is already an interface that tries to compensate the limit of the game. So, while the characters in that precise situation are supposed to communicate with their voice, in our representation we simulate it through text. Teamspeak here replaces the rudimental UI with a direct element. With a more faithful simulation.
The removal of the HUD isn't to hide informations to the players. It isn't to obscure the mechanics to encourage the roleplay as it happened in some games. It's instead the result of delivering the same informations in their natural form.
The ideal game here is the game that doesn't need any specific knowledge. Today the games are complex and aren't that easy to play if you don't have already a background as a gamer. They have their own culture and you need to be dedicate enough to it in order to get involved and have fun. So games in general need study and practice. These are HUGE accessibility problems. Firstly because only a small number of people share those competencies and secondly because the experiences they deliver are too detached and artificial to be enjoyed to their real potential.
The television and cinema are way more direct. In fact they are able to reach a broader public. Yes, you still need the specific competence to understand the cinematographic language but is simpler (and way more common) because what you see on the screen is "analogic" (similar) to your real life experience. So you can relate to a politics discussion, a war, a love story, a fight and so on. You have the competence to understand these stories because you share a common background on this reality.
Games don't offer completely alienated experiences, I wrote about this as well. They still use common experiences like fear, awe, conflict, love, surprise, mystery and so on. All these feelings are again part of a common background. And we definitely do not need to learn "interfaces" to understand them. Everyone in this world, no matter of age and culture, can dream one night to be in a forest and facing a huge orc. And there isn't anything special to learn to relate yourself to something like this.
But if this is a "game", only a very tiny group of people will have the competence to understand the representation filtered through this medium. Because you need to relate to a symbolic representation with its own rules and quirks that only vaguely resembles to the actual situation we want to simulate.
All this is superfluous. It's an useless weight. These are parts that will progressively vanish to favor a more direct experience that doesn't require the player to have already developed specific competencies about "computer games". And it's not just something I say as a personal opinion. But a fact.
If we erase this useless geekdom everyone will be able to relate and enjoy the experiences coming through the games. Hence a broader audience and a better accessibility for the medium as a whole.
This is a very common point of view for the technology in general. The goal has always been about making these tools easier to use so that you don't need a specific knowledge to use them. The technology, in fact, has a barrier to its use represented by the interface. We are moving slowly to remove these interfaces. Till we'll reach a point where the people will be able to "talk" to a machine to have it follow those instructions precisely. Without needing anymore to read manuals, learn icons, buttons etc...
Why? Because the machine itself is finally able to relate to the verbal communication that is what we naturally use (and have competence to).
Take the example of the /wave I made above. If we call a child to watch what is happening on screen when the emote is just a text message, the child will understand what is going on only if he already developed various, external competences. For example he needs to understand what a chat box is, how it behaves. He needs to understand that the the lines he can read are the lines spoken or performed by the characters on screen. All these informations aren't trivial at all. They are for all of us who played these games for so long, but not for him. They aren't so trivial for a broader audience.
Instead. If the model on screen performs the /wave action with a realistic animation, the child won't need any knowledge to understand what is going on. He will smile and tell you directly, "Hey, he's saluting you, look!".
It's not a case that so many people like the emotes and /dance and jump around. These are all "analogic" representations that feel direct. Without the need to parse the game and understand its obscure language.