I trust BJC (referenced in the article), and I also use BJC cables in my systems (though not exclusively).
From the linked article:
Use good quality shielded Cat 6a or better yet Cat 7 cables. Why did I say "good quality"? Read this report from Blue Jeans cable where they tested 20 Category 6 and 6e Ethernet cables only to find that "four met spec, and of those, two did it by a hair." That's an 80% failure rate. If you take issue with a cable company doing the testing, check out what the Fluke Corporation who make Ethernet test devices had to say. Spoiler Alert: Fluke found the same 80% failure rate in Cat6 cables so this is no...fluke. Cables that fail to meet spec, note that every cable from brick-and-mortar stores used in the Blue Jeans test failed, can also inject unwanted noise and errors into your system. You can also avoid this noise issue completely by using optical isolation, i.e. an Ethernet to Fiber media converter, which is something I may explore in the future.
I think the heart of this article was nailed by rojo in his post above. Be sure you get cables that meet spec! It doesn't take a rocket scientist to figure that out!
A couple other comments, they talk about measurements, then slip right back into subjective comments. And, they don't really have any sort of summary / conclusion on the article either! I feel like I want my 5 minutes back after reading that
It's also kind of odd how they talk about noise and it transferring to bits, but you really can't make any general statements like that! You need to know the resolution of a specific DAC to be able to determine if a given amount of noise will make it from the input to the output.
Finally, ONE OF THE BIGGEST ADVANTAGES OF OPERATING IN THE DIGITAL DOMAIN IS THAT IT IS ORDERS OF MAGNITUDE MORE IMMUNE TO NOISE THAN THE ANALOG DOMAIN! An example of this is demonstrated by reading up on Schmitt Triggers
http://en.wikipedia.org/wiki/Schmitt_trigger (not exactly the same, but gets you thinking along the right path when dealing with the digi domain).
