> The author's proposed solution is for sites to use DPI instead, but this would thwart the preferences of users who use scaling because they want the whole site to look bigger.
Agree. I sometimes browse the web on my TV - and I sit about 2 metres (or 6 feet) away from it. So the TV screen occupies a similar portion of my field of view that a phone would, and I usually browse at 150-200% zoom.
DPI might help you figure out the size of the screen, but it doesn't tell you how much of a person's field of view that screen occupies, which is what really matters.
Yes, field of view is what really matters, but scaling at the OS level (like TFA's author is doing) is a good way to correct for unusual fields of view. Got a higher-than-typical DPI and/or ridiculously far away screen and/or vision impairments? Scale the OS up. Got a screen with lower-than-typical DPI and/or touching your nose and/or amazing vision? Scale the OS down. You need the entire OS scaled, not just web pages.
"CSS pixels" are kind of a roundabout way to conceptualize it, but they solve the problem of proper readability once the OS scale is correct for the user. The combination acts as a proxy for field of view.
Then there's detection of course/fine pointing devices, as already discussed, to further augment the layout for proper target area.
I don't believe anything else will matter if the above is accounted for.
Agree. I sometimes browse the web on my TV - and I sit about 2 metres (or 6 feet) away from it. So the TV screen occupies a similar portion of my field of view that a phone would, and I usually browse at 150-200% zoom.
DPI might help you figure out the size of the screen, but it doesn't tell you how much of a person's field of view that screen occupies, which is what really matters.