Anticipating the detection of earths for 3 years is going to be a painful waiting process, but worth it. Meanwhile, it will be fascinating as continued detections(of increasing orbital length) illustrate probabilities of various orbital possibilities. By the way the probability that an "earth" would be in the appropriate geometry to have its transit detected is .5%. So the number of earths detected should be 500xPe. Pe=probability of "earths" around sunlike stars. If Pe is 1%<Pe<10% then that would be 5 to 50 earths detected.

Figuring roughly that among the stars Kepler observed the average HZ 'earth' year is under 10 months lets us know that in this month of data we already 'caught' >10% of the detectable HZ planets in transit. It takes a minimum of 3 observations to confirm a detection but we've already seen a detectible pct of HZ planets, assuming that the total # is significant. Most M star HZ planets should be verified by Thanksgiving at the latest.

Kepler should finally give us enough data to make our first credible models of planetary / solar system formation.

I'm waiting with baited breath to read if Kepler's actual detection sensitivity equals or hopefully exceeds expected design parameters.