I think roleplaying the exact kind of conversation I mention is exceedingly difficult.
The reason I am shocked is because I don't. I once tried to roleplay a game based on Gerrold's "War against the Chtorr". Making the biology of that setting believable to an average nerd is not the same as making that conversation in character believable to my wife with her PhD in biochemistry. If you have some nerd at the table who is playing a character how has IC the same specialized knowledge as they have OOC (a lawyer perhaps playing a character with legal skill) then as GM holding your own in that conversation does require a good deal of OOC prep on your part, but I've rarely had IRL a problem making a conversation believable enough for the players I've had. For one thing, as GM I have full control over the setting and can make crap up to even the playing field, and as long as it feels to the player like a plausible thing with a similar richness of detail they are used to, it works. The goal here is verisimilitude. Plenty of fiction works at that level, even though an expert wouldn't necessarily buy it, the reader has buy in. And really, players are genuinely experts only in narrow areas. It's not that hard to do enough research that they at least enjoy that you've made the effort. If I have a player who is an IRL hostage negotiator, IRL doctor of biology, IRL expert in church Latin, or whatever, I will have a hard time keeping up in their narrow field, but I can fake and the tools for doing that are getting easier all the time.
I am completely aware that groups mange some kind of short-hand here, some of them very effective shorthand.
If by shorthand you mean, "sufficient accuracy for verisimilitude not actual accuracy" then I agree. If you mean that the conversations don't happen, then I don't agree.
I think the various moments of any given conversation are, of themselves, meaningless.
I think they certainly can be. In some systems, and explicitly in many Nar systems, they actually are as often fortune has already been decided and you are just playing out the results of fortune and your acting however engaging doesn't matter, or else it doesn't really matter what you say because the dice are going to decide for you anyway. But I don't personally consider that normal to traditional play or to most systems.
In this post (
Games with great social rules) you described how in a particular system, particular elements of a conversation are used to steer how the ultimate fortune test is resolved. The framing formalized in the system you are describing is something I've been using in even 1e AD&D in a more informal and ad hoc manner (or in the form of predefined minigames written up in the scenario notes) since no later than about 1993 as a means of adjudicating things like bonuses or penalties to Reaction Checks depending on the suitability of the conversation to the NPC's personality and goals. The various moments then in the played out conversation are then extremely meaningful, with different approaches to the conversation leading to very different results. And things like "clocks" are frequently used in an informal way as well, with the conversation perhaps allowing for a natural number of "retries" depending on how the conversation plays out and whether anything truly new is being said, or whether the PCS have become badgering or irritating, repeating the same things or moving on from persuading to some other thing like intimidation or bribery, which tend to on failure result in further conversational options being terminated.
To me this is absolutely normal. Its the way I was taught by example to do it by an older GM in the mid 1980s, and the way I've always experienced at other tables, or when running my own games. I think it is also intuitively normal that they should work that way, as this is how child's games of make-believe work. You actually talk in character.
I'm not at all against formalizing clocks and positioning and difficulty if those things related to spoken things in the conversation. It's good advice and teaches GMs the fundamentals of running engaging social interactions. I've seen some really good advice come out lately on this that is I think much more functional than I find some older attempts that treat social interaction as being a parallel to combat. But I don't want to ever get to the point where the actual spoken words are meaningless.