I think this is wrong because the motivation (as far as we know) for Gygax was to model something, not create a useful or fun gameplay element.
The weapon vs. armor table descends from a similar table in the Chainmail wargame.
The funny thing is the simulation got totally mangled along the way as it was translated to the d20 vs. AC alternate combat system.
The numbers were originally the
target number to kill an opponent wearing that armor with this weapon. But they were turned into
modifiers to the attack vs. AC roll in Greyhawk/1e.
So the mostly flat row of 0s for the Mace originally meant that weapon was
equally lethal against any armor (bit overkill for realism IMO). A very different simulation story from what it becomes in Greyhawk/1e -- no modifiers to
the default sequence of to-hit probabilities vs. AC.
To me the fact that Gygax didn't care about this indicates that simulation was not his main concern. I think he noticed what I noticed using this table in 1e, which is that it makes weapon selection more interesting for the players.
If he wanted the original simulation from Chainmail and made an honest math mistake here, but
never noticed the difference in play, it at least supports my second point that sim rules that play with probabilities are
way less impactful than sim rules that introduce new events into the fiction.