11-16-15: GeorgelofiDepending on the specific designs, though, a sonic advantage may very conceivably result from separating the two functions into separate components, due to the reduction in electrical noise that in an integrated component may couple from the transport section to the D/A converter circuit, resulting in increased jitter (i.e., random or pseudo-random fluctuations in the timing of D/A conversion).
The advantage to having separate transport and dac, is that you can change the dac when you want a change in sound, and you can have multiple sources/transports if your dac has multiple inputs. There is no sound advantage.
On the other hand, though, dividing the two functions between separate components can also result in increased jitter, as a result of various interface-related effects. Including ground loop effects, less than perfect impedance matches, cable effects as George mentioned, jitter in the timing of the S/PDIF or AES/EBU signal supplied by the transport, and jitter that may be introduced by the DAC as it extracts the clock from those signals (if, as is often the case, it is that extracted clock which ultimately controls the timing of D/A conversion).
The bottom line, as is often the case in audio: There are numerous tradeoffs involved, and the net result of those tradeoffs depends on the specific designs.
Regards,
-- Al