I would argue that Artillery (HE) is designed to kill whatever it hits, if not with one round, than through; weight of volley, selection of appropriate weight of shell, or gradual degradation through repetition.
As the most likely target profiles have increased their armor over time, the standard shell weight of what is considered appropriate close support artillery has increased as well (~105mm largely replaced by ~155mm).
The apparent efficiency of the lowest hanging fruit, suppression, is just the obvious acknowlegement of the outsized psychological impact of a shell beyond its physical destructive capacity.
Obviously a shell without the ability to course correct is beholden to the target's decision to remain where the shell has been told to go.
DPICM is not designed to 'destroy that which HE cannot' but instead increase efficiency by reducing effects bleed [inherent through poor TLE and CEP] and a given target's dispersion, and freedom to increase the space, through movement, between itself and the round's point of impact.
HE overpressure at the point of impact is great enough to destroy AFVs, bunkers, buildings. The destructive capacity drops off too quickly however to efficiently compensate for a target's ability to displace or spread out (an armor platoon can present itself as a 50 meter linear target or 300 meter target).
DPICM addresses the problems of dispersion and space between target and point of impact by increasing the surface area of the effect, forcing the target to increase the space between itself and the impact; beyond what is hopefully considered efficient or is immediately prudent for the target.
If you can guarantee a target will be within an HE beaten zone at the time of impact, a probability of effect can be calculated and expected up to and including destruction.