The clothing studio's pod.html compares a factory run with print-on-demand under Poisson demand. I fed its engine odd inputs. A minimum order of 1,000,000 pieces on demand of 100 took 11.7 seconds, long enough to freeze a browser tab.
The cause: the demand table was sized to the run (max(Q, mean + 12 sd + 30)), and the crossover search repeats the calculation 800 times. Past mean + 12 standard deviations the Poisson tail is nil, so a bigger table adds nothing. Sizing it to that bound alone gives the same figures (all 12 tests still pass) and the million-piece case now takes about 0.2 seconds.
A second, smaller defect: passing podPrice: undefined explicitly was refused instead of meaning "same as the shelf price". Undefined fields are now ignored.
Limits: the inputs were invented placeholders, and no real designer's numbers have been run through this page.