I always fill the pan - ignition as fast as possible.
In Pletch's testing, holes at the bottom of the pan, not at the sunset location, were the fastest,
so the priming burning down to the hole does not slow down ignition. The pan goes off instantly
and it's the heat that sets off the main charge with a gun having a good liner. Now, if there is a long
channel to the power, then ignition will be slowed if the powder must burn to the main charge. This
makes sense.
Seems to me, that the testing I have read about, unconfined black powder burns at a rate of around 1,000fps, or thereabouts.
I do not know what granulation "they" were referring to.