I'm waiting for the super happy people to bust out Unlimited Blade Works, any time now.

The distinction between kiritsugu and Confessor is quite interesting. It hadn't quite occurred to me that there could be distinct branches of rationalism. That's to say, I knew that two rational agents could disagree about something -- that's a consequence of Bayes -- but it hadn't quite occurred to me that the space of rational agents was quite that large.

I imagine you've covered some of this in your work on metaethics; I should probably dig that up.