This ^ is just made up bullsh1t.
To get sufficient noise on the cable to impact electronics on either end, you'd be completely destroying the data going on the wire. Unless you're placing a running a vacuum cleaner and hair drier on top of the ethernet cable, you're not experiencing this.
And if you have a DAC with a crappy enough design (a single packet sized buffer?) that it could be effected by a dropped (and then automatically re-sent) ethernet packet, then seriously, W.T.F?!?
If it wasn't a reproducible, double-blind test, then it's just the placebo effect.
There's nothing wrong with the placebo effect. It's an awesome effect. If it makes your enjoyment of audio better, then by all means, turn up the placebo! I personally rely on the placebo effect all of the time, and have no shame in doing so. I spend good money on AV placebo equipment. And I love it.
But I am careful not to mislead others about the difference between reality and placebo.
We seem to have different understandings of the word "theoretical".
On the one hand, we have actual, measured, tested, real-world proven, fifty years worth of proof that basic datagram packets over Ethernet could experience an error rate of statistical zero (not actual zero, but close enough that you are unlikely to experience a single such error in your lifetime). And that's before you add on any high level protocols that can easily provide an actual zero error rate.
On the other hand, we have people making ridiculous, anti-reality claims, backed up solely by "I may sell these expensive cables to make money, but trust me, I can hear the difference!"
Yes. Digital data is "just electric [sic] voltage signals". Sure, a few trillion dollars has been made by figuring out how to correctly deal with those "just electric [sic] voltage signals". Minor little start-ups like Intel, ARM, Asus, AMD, Fairchild, Apple, AT&T, Acer, IBM, TSMC, Samsung, Honeywell, Western Digital, SK Hynix, ...
Dealing with those signals let us go to the moon. Allowed us to decode the human genome. Let's us fairly accurately predict weather. Take pictures of distant galaxies. Carry an always-connected combination GPS + computer + telephony device in your pocket. Little things like that.
As an audio equipment maker, you'd have to be pretty stupid to route noise from the PHY of the NIC into your amplifier, but I guess that is possible to achieve, if you wanted to. However, I'd suggest avoiding audio equipment from anyone who designs their equipment to route noise from the PHY of the NIC into the analog output from the DAC.
And the article that you linked to has nothing to do with Ethernet.