We have to be really really careful here, and Paul isn’t. You can’t conflate these domains just because they’re digital.
Streaming consists of pre- and post-streamer; the former (ethernet) is asynchronous and jitter/timing is simply not a thing - by definition. Packets are sent but are also resent if they are checked and found wanting. Timing comes in after the streamer and post-streamer is where reclocking makes a difference, etc.
"It strips away the noise and aligns every packet so it arrives precisely when it should". The first part of this is great but the second demonstrates a complete misunderstanding of what in the network domain can and can’t affect sound quality.