Knowing About Biases Can Hurt People

Um, really? The primary point there point there doesn't seem to be about whether only the in-crowd should know about things but whether knowing about cognitive biases is safe for anyone. But maybe that's just me.

many of the comments in Crises of Faith

But that doesn't fit with your claim that the truth is being reserved for some elect in-group. No one (from my quick reading of the thread) is talking about just telling things to people who are already members of some super-secret transhumanist rationalist amalgam, just that there are some people out there who aren't going to respond productively to being told certain things. So the claimed in-group includes a large number of people who haven't ever heard of Less Wrong.

Regarding the issues with gnosticism, that's not an accurate summary of gnosticism in that gnostics thought that there was a God superior to the Demiurge. This seems pretty different from a view which has no deity at all and a plan to construct something that if you squint at it in a funny way might sort of resemble a deity if certain results occur. And no one, absolutely no one, would claim that we somehow have to rediscover pre-existing knowledge to be more rational to help make Friendly AI. Indeed, Eliezer has repeatedly emphasized that an important part of science and actual productivity is to realize that there aren't any ancient sources of hidden knowledge, and that a logician today is really much more worth listening to than anything in Aristotle. (See, e.g. his comments here).

don't normally even go together and putting them all together as an attempt to make something which (vaguely) resembles a religion.

They are all here on this site.

So the argument is that they are all mentioned here and that some people subscribe to some or all of them? Sure, and some people here like the Red Sox. And a lot of others think that studying pure math is fun. And we have a sizable fraction that thinks that everyone should know how to use Spivak pronons. The presence of certain classes of opinions in some fraction of a population doesn't necessarily tell you very much.