Digital is an RF signal and changing cable length with an RF signal WILL alter the response of the circuit. This is true UNLESS the input, output and cable impedance are all identical. Anybody that tells you differently lacks experience in this specific field, regardless of the cost of their cabling or their past credentials.
Having said that, the wavelength at these frequencies is VERY long. I don't know how much of a difference such a short difference in cable length would alter actual impedances in a less than optimally matched sytem, but it has been shown that it will alter the return time of the reflections that occur. Depending on the stability of the output stage in the transport, it might be important or it might be moot.
If someone REALLY wants to know, i can do some impromptu testing and provide "less than set in stone" type results using a digital analyzer and a make-shift circuit. I can take some 75 ohm coaxial stubs of 1 and 1.5 meter length and measure the feedpoint impedance when shunted. Only problem is that all of my test gear is set up for a 50 ohm nominal impedance, so the results may not be exactly perfect. It would give us an idea as to whether or not there was a sizable shift in loading conditions though.
With all of that in mind, if one terminates the cable with the appropriate value resistor at the input of the DAC, most of this becomes moot. The value of the resistor used would vary depending on the actual components and cable completing the circuit. The first company that i know of to take this approach was Tara Labs and that was so long ago that Tara didn't even remember doing it. Sean