They're worried about the idiot that builds a
one-off programmer for use
in production, and doesn't test it thoroughly, and then destroys
hundreds or thousands of chips.
How does this differ from the ifiot who designes a PSU which puts a 25V
spike on the 5V rail at power up/power down (and yse I have seen a
regualtor circuit which does that -- in a commercial bench PSU!) and then
tries to cleim all your ICs were defective? Or the idiot who takes zero
anti-static precautions and then claims yuor ICs are failing i nthe field
a couple of months later?
-tony