I don't think it's accurate though. The magnetic field is radial (the flux lines all point inward to a central point), there is no force exerted by the magnetic field along the lines of flux. Force is always perpendicular to the flux.
What does happen is the vibration will tend to allow all radial forces exerted by the spider and surround to be equalised. Any slight unbalanced force caused by a skew surround will be stressed further, and IF there is a weak spot such as still-liquid glue, the structure will physically move to minimise that force.
You are correct regarding the flux field produced by the magnet; it is static,with the lines encircling the voice coil gap as illustrated in this Klipsch low frequency driver flux diagram:
However, the field produced around the voice coil is dynamic in response to the current it carries. This current varies, changing the voice coil field, which then interacts with the static field in the voice coil gap. Since we are injecting an arbitrary frequency of, say 100hz, then the field will grow as the signal amplitude rises to peak. At peak, the voice coil field will be repelled out of the gap, one direction or the other, depending on polarity, by the static gap field.
As the signal passes peak and heads towards zero, the voice coil field collapses, allowing the voice coil to relax back to it's neutral position within the gap. The process repeats itself as the sine wave of the signal heads toward peak at the opposite polarity, deflecting the voice coil in the other direction.
The voice coil field is actually oscillating (maybe pulsing might provide a better illustration) in response to the 100hz sine wave we are injecting.
I hope this helps and doesn't sidetrack the op's original intent for this thread. This stuff is fascinating to me and I appreciate all the insight and sharing.