That's not the first time, nor will probably not be the last time, a reviewer will make that comment. Their confusion is understandable since they assume a TL can only be one with a really long line, with a hole in the wall for the terminus, whose 1/4-wavelength resonant frequency of its length is the sole determiner of the shape of the system response curve and the system's overall tuning frequency. They also are bothered by there being a port tube and the impedance curve having two peaks, just like a vented system. Of course they either don't know or won't admit that ALL TLs will have twin impedanc peaks until stuffing is added, meaning TLs are inherently 4th-order systems just like vented system. The SongTowers as an ML-TL uses the internal height of the cabinet as the line's length. That length is long enough to allow its 1/4-wavelength resonant frequency to contribute to part of the system's overall tuning frequency, and the rest is contributed by the mass-loading port's dimensions. And, the port's specific location is chosen to optimize the system response. It's quite easy to demonstrate that an ML-TL is not "just" a vented system via modeling its response. Theoretically, if it's just a vented system, one could cut the cabinet height in half, double the cabinet depth to maintain the exact same cabinet volume, and keep the same port dimensions, resulting in the system's tuning frequency and overall system response not changing. It won't happen! In the end it comes down to mostly semantics, definitions and perceptions, but that doesn't change this fact: the SongTowers perform extremely well.
Paul (designer of the SongTowers' ML-TL)