Personally, the idea of the crisis of faith seems to me like a symptom of bad thinking. If you hold an idea so tight that you need a crisis of faith, then maybe the real problem is that you have any belief held so strongly that it takes that much effort to overcome it. Aside from human-level stupidity in-the-moment of discussion (e.g. supporting a position despite contrary evidence because it's the side you picked at the start), a master rationalist shouldn't need to have a crisis of faith.
I had the last thing that felt like a crisis of faith to me when I was 18. Since then I haven't been able to hold onto any belief so tightly that any normal amount of rationality effort hasn't been able to change my mind.
That would be all very well for master rationalists so expert that they have no beliefs that might require a crisis of faith. I don't happen to know any of those; do you? I would be skeptical about anyone (yourself included; I hope you aren't offended) who claims to have none: how do you know you aren't merely failing to notice some tightly-held beliefs?
Since there's been much questioning of late over "What good is advanced rationality in the real world?", I'd like to remind everyone that it isn't all about post-doctoral-level reductionism.
In particular, as a technique that seems like it ought to be useful in the real world, I exhibit the highly advanced, difficult, multi-component Crisis of Faith aka Reacting To The Damn Evidence aka Actually Changing Your Mind.
Scanning through this post and the list of sub-posts at the bottom (EDIT: copied to below the fold) should certainly qualify it as "extreme rationality" or "advanced rationality" or "x-rationality" or "Bayescraft" or whatever you want to distinguish from "traditional rationality as passed down from Richard Feynman".
An actual sit-down-for-an-hour Crisis of Faith might be something you'd only use once or twice in every year or two, but on important occasions. And the components are often things that you could practice day in and day out, also to positive effect.
I think this is the strongest foot that I could put forward for "real-world" uses of my essays. (Anyone care to nominate an alternative?)
Below the fold, I copy and paste the list of components from the original post, so that we have them at hand: