The coil works as a dynamic resistance or impedance if you like.  The
secondary DC resistance of 7.5K has virtually no effect on the primary
current draw under steady state (DC) operation.

The basics are:   Run current through the primary until the core is
"saturated" then abruptly stop the current flow in the primary.   The field
induced by the saturated core collapses and causes a high voltage spike to
occur on the secondary lead to the distributor.  This spike  >20KV is timed
to correspond to the juxtaposition of the rotor and one (or more) of the
distributor contacts causing a >20KV "spark" or arc at the spark plug gap.
Mr. Lenz said it well with the voltage across a "coil" being proportional to
the change in current.  With an "infinite" change in current over time one
would expect a large voltage change across the output terminals.

From Kettering..."the only parts with an infinite service life are those
which are never installed"  Kettering invented this type of ignition system
and it bears his name.

So a primary fire cound be  a DC issue with a short to ground causing the
points when closed to fuse and direct battery current through the primary of
the coil.  Measure the DC primary resistance.  Divide 12VDC by this
resistance and you will calculate the current through the primary.   It has
been my experience that you can short the primary to ground with 12VDC
applied without inflammatory results.  But hey, I'm just a dwEEb.

A better answer might be a primary feed wire grounding without going through
the primary coil.  This would seriously smoke the insulation on said wire
and maybe with enough inflammable grunge on the wiring cause a "fire".

Donning my nomex and awaiting the flames ....

pensioner