@mahler123 - you said:
Essentially what I see in your post is the buffering of streaming content. Am I off base here?
That's half of the equation, yes. The Diretta Host computer buffers streaming content as you have described. I've observed up to eight seconds.
To build that buffer, it must receive audio data much faster than real time and typically in packet bursts. For the Diretta design, this is has no negative consequences to audio quality because filling the buffer happens on a computer that's not connected to the DAC.
The Diretta Host, then, carefully meters out the audio data in small, precisely timed transmissions over a private point-to-point link to the Diretta Target. The Target receives audio data "just-in-time" with minimal (virtually no) buffering. Its job is simply to copy audio data from the network interface to the a USB DAC or DDC.
Network isolation and smooth data delivery with minimal synchronization overhead are what sets this "three-tier" architecture apart.

