I've just perused the article in the Australian HiFi Review.
On page 82 with the impedance curves, IMO, there's has to be an error, and most probably more than one. They show a curve with both ports open, one with one bung blocking one of the ports and one with both ports blocked:
1. By the addition of one bung, the box resonance frequency has to increase but it's very similar on the graph
2. If you use the two bungs, the box resonance will grossly increase in frequency. Also, if the two ports are blocked, the box for practical purposes acts as a sealed box and therefore, there should be only one impedance peak not two!
I don't trust the information in that review.