> As much as I'm a fan of Gwern's work the conclusion he reaches is dead wrong. A high rate of attrition does not prove that brain washing does not work

Yes, it does. What else could possibly show it doesn't work better than the fact that it doesn't work?