The precise shape of the speaker has clearly been designed to work against a wall. Look at the planform. There is intentional diffraction designed for midrange and there is going to be some element of horn loading. Without the rear wall the speaker cannot operate as designed. That is pretty clear.
Think about the wavelengths. This is key. 200Hz is 1.5m. 600Hz is 0.5m The speaker is large wrt these wavelengths, but not huge. 0.5m is going to result is some significant path length differences. Half space radiation versus full space is going to make a big difference too. The woofer is likely significantly loaded when against the wall, so in free space the bass alignment will be well out. At less than about 1kHz (0.3m wavelength) there is going to be no reflected sound from the woofer. It is acting as a horn loaded driver diffracting out past the edges of the cabinet.
Second guessing the exact design reasoning is going to be futile absent a reasonably serious simulation and knowledge of the driver parameters. Given the pedigree of the speaker one might guess that the designers have a bit more of a clue than they are being given credit for here.
I imagine Klippel have a solution to measuring surface mounted speakers, but as usual, I bet it is an optional component that requires one's first born child in payment.