The author seems to struggle with the concepts.

Confusing storage bit rot with memory corruption.

It gets worse from there....

Author seems unaware all Intel processors at the hardware level support ECC(yet is soft disabled), it's just not enabled for consumer products at the driver or firmware level. Intel used to sell people people "hardware upgrades", which were really software upgrades that merely enabled the functionality of the existing hardware, Hyper threading on lower end i3 and i5 processors. Also enablement of on die cache that was fully functional but just disabled by deliberate out of box configuration.

Author also then glosses over the really sad part that ECC demands a premium not due to better hardware, but because it's simply more stable. It's mediocre RAM with basically just a configuration file that says do ECC(it's otherwise basically identical). It's really nothing that special besides running additional traces in the motherboard. That's where Intel saved "big money" by not supporting ECC on consumer chipsets. Motherboard design is somewhat simplified by a very small margin.

This isn't price targeting as much as it is market segmentation and product differentiation and just being cheap (profit driven).

Server operators would still use server grade equipment for numerous other features(like multi socket boards, fat PCI buses/lanes, massive amounts of RAM potential, to name a few). To simply call ECC that which makes people by Xeons is poorly misguided.

I think the better term would be price discrimination:

https://en.m.wikipedia.org/wiki/Price_discrimination