Not to be comparing oranges and bananas, but it has always been easier to create something good in the digital spectrum of things than something good in the analogue domain. That is also the reason why it is possible to make a cheap but good quality D-A but also A-D converters. However, if you want to go above the current standards of what OEM industry can provide on a daily basis, then you need funds and as a result, high end digital technology is expensive. While I have no doubt the quality of the converter in the HDVD800 is good, it is also true that it must have taken them a lot more time and to perfect the analogue circuitry of the amplifier.
What I don't understand is why some people insist on the output impedance being as low as possible and wave with the figure like with a magic wand. It says nothing about the way the amplifier is constructed. What matters is how the output impedance has been achieved and what compromises were made in order to achieve it. What good is an amplifier that has a 1R resistance when the manufacturer used a massive amount of feedback to accomplish it? Soecification is a specification and in some cases it tells you something and in others it tells you nothing. It never tells you everything.
Like I said in the SPL Auditor review, it is clear that that amplifier was designed with minimum compromise in mind so that in the end, the impedance was what it was, 9R if I remember correctly. It might translate to some hum with lower impedance headphones but I'd rather have that than completely dampened dynamics and colouration. I don't own the Auditor, this is just an example.
Also, the crosstalk attenuation between the channes is usually completely misunderstood and people usually translate it to wider sound stage. That is only partially true because crosstalk is not a constant figure but is frequency-dependent. So, if Sennheiser quoted it to be 93dB measured at 1 kHz, it will undoubtedly be a lot lower at 20 kHz.
Listen the amplifier for yourself, don't believe everything they are saying.