Yes but keep in mind that due to how trajectories are determined, the probability will keep going up until it suddenly drops to zero once its refined enough (i.e. the cone of futures can shrink while still including Earth - just less of Earth - for a while before suddenly it's accurate enough that it doesn't).

> We also must consider that the likelihood for this specific event has been increasing, no one says this is as high as it goes.

As the measurement of the asteroid's position and velocity becomes more accurate, it's going to increase (still possible with smaller error bars on the measurement) or be set to zero.