Discussion in comments below this answer relate to how much solid rockets or SRBs burn bottom-to-top versus radially outward.
I am sure I've seen a diagram somewhere showing that different designs are different in this aspect, and that in some cases there was an advantage for the lower part to burn radially faster than the top in order to allow for "better flow" but I really can't remember exactly how the discussion went.
Am I imagining things and all SRBs quickly start burning throughout the length, and the burn rate is uniform along the whole length and strictly radial? Or are there some designs where the mass at the bottom will disappear faster than that at the top, resulting in at least a small but significant translation upward of the center-of-mass?
If this does happen, it would be great to have an example or extreme case where this happens more than for typical SRB burn patterns.