Their way of specifying the test condition is a bit shady, IMHO. Measuring THD+N with A-weighting? Yeah, the power-line related noise would be conveniently attenuated, so are the higher order harmonics. That could also explain the interesting looking THD+N vs frequency curve, which is pretty much the shape of an inverted A-weighting curve. You can argue 'that 's all you can hear', but comparing numbers needs to be done apple-to-apple.
Also, the 'THD+N@20kBW (No-wt)' is questionable. What does that even mean? THD+N @ 1KHz with 20KHz measurement bandwidth, or THD+N @ 20KHz with an unspecified measurement bandwidth? Either way the result is significantly lower compared to a well-built DIY amp such as the KGSSHV Carbon or Blue Hawaii. The FFT and multi-tone plots are the telltale of the amp's linearity.
The 80uV noise is good but not something worth bragging about, given the low gain of the amp.
Frequency response: 20Hz-40kHz (±2.5dB) means the highest and lowest point across the measurement range can be as large as 5dB apart. Go figure.
The interesting dip between 10kHz and 20kHz on the frequency response curve indicates that they need to do a better job calibrating the high voltage attenuator in front of the AP.