One thing I hear you saying here is, "We shouldn't build social institutions and norms on the assumption that members of our in-group are unusually rational." This seems right, and obviously so. We should expect people here to be humans and to have the usual human needs for community, assurance, social pleasantries, and so on; as well as the usual human flaws of defensiveness, in-group biases, self-serving biases, motivated skepticism, and so on.

Putting on the "defensive LW phyggist" hat: Eliezer pointed out a long time ago that knowing about biases can hurt people, and the "clever arguer" is a negative trope throughout that swath of the sequences. The concerns you're raising aren't really news here ...

Taking the hat off again: ... but it's a good idea to remind people of them, anyway!

Regarding jargon: I don't think the "jargon as membership signaling" approach can be taken very far. Sure, signaling is one factor, but there are others, such as —

• Jargon as context marker. By using jargon that we share, I indicate that I will understand references to concepts that we also share. This is distinct from signaling that we are social allies; it tells you what concepts you can expect me to understand.
• Jargon as precision. Communities that talk about a particular topic a lot will develop more fine-grained distinctions about it. In casual conversation, a group of widgets is more-or-less the same as a set of widgets; but to a mathematician, "group" and "set" refer to distinct concepts.
• Jargon as vividness. When a community have vivid stories about a topic, referring to the story can communicate more vividly than merely mentioning the topic. Dropping a Hamlet reference can more vividly convey indecisiveness than merely saying "I am indecisive."