If I'm understanding your comment correctly, I strongly disagree with this way of framing such suggestions. It seems anathema to the rationalist enterprise. Many rationalist simplifications of or modifications to (social) interaction, or other not-strictly-rationalist approaches that are regardless endorsed by us, are hit by your argument. E.g. requesting tabooing words, requesting predictions of differing anticipated experiences, Crocker's rules, confessing noticing confusion, etc. etc. on through the Sequences et al.

A core of the rationalist ideal is to take approaches that promote the discovery, recognition, and sharing of truth except where there are situational reasons to hold off on doing so in those specific cases. For example, I agree with warnings that have been raised in the comments on this post about trying Telling without a cooperating or rationalist receiver. But that's in the same way that asking a Muggle to taboo their words can be a not-so-great idea.

I suspect that high-profile Bay Area (and possibly New York?) rationalists would bear this out. As a specific example, as far as I can tell, Alicorn seems to be the rationalist master of Telling and generally avoiding beating about the bush when she wants something, and wins because of it. More generally, from what I gather as a spectator, there seem to be a lot of techniques or behaviours on instrumental, emotional, and interpersonal fronts that are making the Bay Area awesome and an ever-stronger attractor to rationalists around the world, but which the broader rationalist/LW community does not necessarily hear about.

The success of the Bay Area subcommunity's approach seems somewhat unknown. And I think that means that when someone comes along from there and says to the broader community, 'Hey, we should try Telling more,' there is a lot of cultural context (of the Bay Area generally, and all the interrelations with communication systems, openness, etc.), experience, and success underlying that suggestion that is not visible. I think if enough commenters adopted this approach, it would becomes recognised, not be misinterpreted, and work. Now Brienne's posted this, possibly even people can link to this post to try to prevent being misinterpreted when they are Telling on LW.

A lot of the Bay Area's success seems to come from people taking simplifying approaches to communication seriously and cooperating. When you say

It's also not so obvious that you can effectively change conventions like these by just starting in and asking others to change. If you tried your "developing trust" tactic with me, I'd probably play along to avoid conflict on one occasion, and avoid YOU after that.

that pretty much feels like the complete opposite, i.e. writing off the suggestion and anyone who takes it seriously. I'm not sure if I'd call it defection, but it has a similar feel. On a collective level, both the receptive and the skeptical attitudes are self-fulfilling, because these kinds of things really do seem to work when enough people take them seriously, and will certainly fail if everyone scorns them. (E.g. look at how many memes from the Sequences are pretty much unanimously taken seriously.)

(I acknowledge that I might have completely misread your comment.)