Nitpick: while a significant fraction of rational people are not polyamorous, polyamory could still be better evidence for rationality than atheism. That's because there is so much atheists around, many of which became atheists for the wrong reasons (being raised as such, rebellion…).

Let's try some math with a false dichotomy approximation: someone could be Rational (or not), pOlyamorous (or not), and Atheist (or not). We want to measure how much evidence pOlyamory and Atheism are evidence for Rationality, given Background information B. Those are:

• Atheism gives 10 log(P(A|RB)÷P(A|¬RB)) decibels of evidence for Rationality
• pOlyamory gives 10 log(P(O|RB)÷P(O|¬RB)) decibels of evidence for rationality

Now imagine that B tells the following: "Among the 6 billion people on Earth, about 1 billion are atheists, 10 millions are rational, and 1 million is polyamorous. Every rational people are atheists, and 5% of them are polyamorous".

So:

• P(A|RB) = 1
• P(A|¬RB) = 990,000,000÷5,990,000,000 = 99÷599 ~= 0.17
• P(O|RB) = 0.05
• P(O|¬RB)) = 500,000÷5,999,500,000 = 5÷59995 ~= 8.3×10⁻⁵

Applying the two formulas above, Atheism gives about 8 decibels of evidence for rationality. Polyamory on the other hand, gives about 28. And rationality itself, P(R|B), starts at about -28. pOlyamory is enough to seriously doubt the irRationality of someone, while Atheism doesn't even raise it above the "should think about it" threshold.

If this is not intuitive, keep in mind that according to B, only 1% of Atheists are Rational, while a whooping 50% of pOlyamorous people are. Well, with those made up numbers anyway. Real numbers are most probably less extreme than that. But I still expect to find more rationalists among a polyamorous sample than among an atheist sample.