It's possible, but we're taking a lot of this "study" on blind faith from Athiest Shoes. They didn't explicitly spell out their methodology. I find a sample size of 168 packages pretty small given the volume of US mail there can be.
Once scenario is that someone saw the tape and used that as a visual reference to organize the packages. If all of those were sent out as a group in a single shipment, that would skew results to appear that the labeling causes the delay (technically, yes, it did, but not because of some misguided religious notion). The other packages were simply not labeled at all, and wouldn't have been lumped into a single group using the same visual indicator (the tape) above.
So package group A (with the tape) would act as a single shipment, and a single delay causes a delay for everything. Package group B gets dispersed among many shipments, allowing more tolerance for delays - a single delayed shipment affects a smaller percentage of Group B.
We could sit here and pick apart each others' analysis, but the bottom line is we're looking at a pretty vague inforgraphic and extrapolating some pretty sweeping generalizations from it.