The first number you give is wildly inaccurate. It rises to about 35% in the +/-1 day window around ovulation, with lower probability during a wider weeklong window, and very near 0 for the rest of the menstrual cycle. I'm fairly sure the actual compounded odds of conception from procreative sex at a random point in the woman's cycle is much closer to 1 in 1000.

Also unprotected sex does not necessarily equal procreative sex, which I realize is a somewhat controversial statement, but nonetheless true.