The author's argument for high dpi is unconvincing. I mean, if you magnify something by 24x of course it's going to look blurry. That doesn't mean that you actually need/want that high pixel density. You can use the same argument technique to argue for 8k/16k displays. It's also strange how the author mentions that cleartype is a thing, but for the conclusion comparison he ignores it entirely. If anything, by taking cleartype rendered text and magnifying it (thereby making subpixel rendering pointless), he's making low-DPI look artificially worse.
He also makes the baffling suggestion to use 2x scaling for 4k. That means that your 4K monitor has the same amount of screen real estate as a 1080p monitor, a reduction in surface area of 43% compared to a sane low-dpi monitor @ 1440p. While I agree that it would likely result in better looking scaling, I can't imagine how slightly better looking text/scaling can make up for the loss in surface area.
Personally I've regularly used 4k 27" and 1440p 27" monitors, and while the 4K monitor is clearer, it's not something that I'd fork over a few hundred dollars for.
He also makes the baffling suggestion to use 2x scaling for 4k. That means that your 4K monitor has the same amount of screen real estate as a 1080p monitor, a reduction in surface area of 43% compared to a sane low-dpi monitor @ 1440p. While I agree that it would likely result in better looking scaling, I can't imagine how slightly better looking text/scaling can make up for the loss in surface area.
Personally I've regularly used 4k 27" and 1440p 27" monitors, and while the 4K monitor is clearer, it's not something that I'd fork over a few hundred dollars for.