Humans are built to argue and persuade. We tend to win when we endorse arguments that others accept, and win even more when we can generate new arguments that others will accept. This is both because people notice who originated the arguments that they accept, and because this ability helps us to move others toward opinions that favor our policies and people.
Very interesting framing of classes of arguments. There is another dynamic I think might be worth identifying in this context, which is when a less-informed arguer attempts to present an argument to find why experts disagree, or where they are wrong.
It's especially important when it allows the people below that 80% threshold to get closer - a mechanism that internet conversations was once thought of facilitating. It turns out that the more common dynamic is... less inspiring.
Isn't the problem that when you are more capable than your audience they usually parrot an "expert" opinion? If you try to convince them of otherwise it becomes too complex for them and they just rethrow the opinion at you again feeling they are right because you failed to convince them. This is one of the reasons whole institutions and norms in society can exist even though they are severely flawed.
My strategy claims were relative to the goals I identified.
Re "These three cases illustrate the three general cases, where your main audience is 1) about as capable , 2) more capable, or 3) less capable than you in generating and evaluating arguments on this topic. Your optimal argumentation strategy depends on in which of these cases you find yourself."
Suppose the audience and you have different opinions of capability. It seems to me that the optimal argumentation strategy and hence value of the conversation is is a function of the degree of agreement between the two perspectives.
Furthermore, the strategy is adjusted by the relative power (i.e., power to compel an outcome) of the audience and you.
When your audience is much less capable than you, then arguments that persuade you tend to be too complex to persuade them. So you must instead search for arguments that will persuade them, even if they seem wrong to you. That is, you must pander.
Must? You're saying then that intellectual dishonesty is mandatory when addressing epistemic inferiors? Then you must pander (search for arguments you know lack merit) to your students?
Your definition of pandering as searching for arguments you know lack merit is incisive, but do you really mean to say pandering is unavoidable? (The honest alternative, it seems to me, is actually a kind of parroting - an appeal to authorities.)
You should have a name for that audience - eg "the reference audience". If it's right that the reference audience is substantially below the very best that could mean that the returns to greater skill diminish once you've passed the reference audience. The very best can come up with arguments which are too complex for the reference audience. Unless they restrain themselves from giving those arguments, that ability could have disadvantageous effects.
Somewhat overlooks other methods of dealing with arguments, changing the subject, casting doubt, privileged knowledge, us against them, avoiding them or making them beside the point.
An extreme case of "parroting" what experts say happens online, though the sharing of links. This is lower effort than writing things yourself, and often more useful, since the reader (who may know more than you) can go to the source and judge for themselves.
As a sharer of links, it's somewhat harder to make an embarrassing mistake; you're judged mostly by the quality of your links, which means judging the argument's plausibility, both to you and to your audience. Selectively quoting from the article tends to be pretty safe too. It can be difficult to do a better job at explanation than a skilled writer, and as you say, paraphrasing an argument risks introducing an error, particularly when moving quickly.
In conversation, the equivalent might be to say "I don't know, ask Google" or if you're feeling generous, do the search yourself and have a conversation about what you find. In a conversation of which route to take, checking traffic on your phone will help you make a compelling argument.
With link sharing, perhaps there are fewer opportunities for demonstrating intelligence, but more opportunities for showing good judgement, since you're judged by the relevance and plausibility of what you share.
The downsides are lack of originality and the likelihood of repeating misinformation when your judgement fails. But, for *most* subjects beyond those you have personal experience with (which is after all, most of what's going on in the world), it's probably a better bet.
Perhaps younger generations will find most conversations without technical assistance to be ignorant, vague, and unreliable? The denigration of "mansplaining", while mostly associated with feminism, might also reflect a growing disdain for speculative theorizing or philosophizing.
But I expect there will still be plenty of places on the Internet open to such things.