Strict honest Bayesians? ZERO. (Not even LW contains a single true honest Bayesian.)

Approximations of honest Bayesians? Better than you might think. Certainly LW is full of reasonably good approximations, and in studies about 80% of people are honest (though most people assume that only 50% of people are honest, a phenomenon known as the Trust Gap). The Bayesian part is harder, since people who are say, religious, or superstitious, or believe in various other obviously false things, clearly don't qualify.