Wow, nice start! Unfortunately none of your citations, except maybe those behind paywalls, seem to address our main issue directly. They do, however, provide much useful background about the emotional and interpersonal dimensions of persuasion and deconversion. Streib's book, for example, uses personality tests (the 5-factor measure, a measure of "well-being and growth", and such) to study deconversion. Jacobs writes that "severing of socio-emotional bonds to the religious leader" is important in deconversion. And Ullman writes that "Emotional factors were more closely associated with religious conversion" than cognitive personality features like tolerance of ambiguity. Max Heirich disputes the emotions-and-relationships focus however.

Cobb and Kuklinski do examine argument styles for effectiveness, but the only dimensions they study are Pro/Con and Easy/Hard (to comprehend). From other browsing, I got the impression that Hard to Comprehend means that the arguer actually explains how the policy will lead to great or terrible things, rather than simply asserting that it will. Ah, democracy!

It occurred to me that political argument is probably better studied than religious argument, and is similar enough, dynamically, that an argument tone effective in one area is probably comparably effective in the other. That search led me to:

• Arceneaux, Kevin. 2012. “Cognitive Biases and the Strength of Political Arguments.” American Journal of Political Science, 56 (2): 271-85.
• JamesCroft, 2012, "The Freethinkers’ Political Textbook – Logos, Ethos, Pathos"

Arceneaux finds that exploiting in-group bias and loss aversion make for effective political arguments. Croft cites psychologist and political consultant Drew Westen for the claim, among others, that tone matters, and that in particular, condescension or know-it-all attitudes are trouble. However, Croft does not cite specific studies or explain any findings.