BTW, the PD design has space for 3 large input caps, but only 2 are mounted. If a third is added, the capacitors will charge faster and hold more during the short period of time the peak of the input AC wave form is above the critical value. It could reduce the PD's sensitivity to low input AC voltage.
I'm tempted to add back the third cap (which I'm sure was removed to save pennies) but I need to calculate the inrush current to make sure I won't burn anything out. After I nail down the sensitivity with the current design and I'm back home, I may add back that 3rd cap (it's shown in the patent also) to see if it improves things.