Hi two cents -
Here are some thought (but no answers) on your questions:

1) Interference - The motion on the seismograms is long-period wave-like motion.
Electrical interference would be short-period jerky - like you see in thunderstorms.
This suggests the motion is real ground motion being recorded.

2) A model in which the atmosphere is compressed by the incoming solar
wind on the side facing the sun and drawn out on the side away from the
sun as the solar wind sweeps atmosphere away (this is the real world effect
of such a storm, well known and observed) might suggest that if the ground
motion is related to the atmospheric compression/rarefaction, different
first motions might be observed on the two hemispheres - sub-solar and
anti-solar.

3) With flares, there is no time delay between observed flare and result because
we observe the flare from observatories on the earth. So our observation
is not the "true" time of the flare, but the time it is observed on the earth -
approximately 8 minutes later.
With geomagnetic storms, the storm is at the earth not the sun. This is
a disturbance of the geomagnetic field of the earth. No time delay would
be expected.