If I gently hit the ball with a moderate amount of brushing to impart a spiral spin, I can clearly see that the ball has a strong component of spiral spin from the onset. The axis of rotation is clearly not perpendicular to the flight of the ball in this case. For those of you still following along -- the axis of rotation of a spinning ball is perpendicular to the direction of its spin. For example, the spin for a pure topspin ball is in the direction of the flight of the ball. However, its axis of rotation is perpendicular to the flight. The same is true for underspin and both flavors of vertical-axis spin (sidespin). OTOH, just the opposite is true for both flavors of spiralspin -- spin is perpendicular to the flight but the axis of rotation is in the same direction as the flight of the ball.
While spiralspin spin was very evident at the onset (coming off the strings) in my experiment, its effect is not manifested until the bounce. Spiral spin is not caused by the bounce -- it is already present and affects the bounce. Gravity is not magically producing spiralspin either -- the spin happens as a direct and immediate effect of the brushing motion.