This article is a bit dangerous because it, like all of Anthony’s previously posted work, lacks any user representation or testing to iron out the assumptions.
For example it’s noted that white offers better legibility than black. Therefore IPSO FATSO WCAG isn’t right take it w a grain of salt. Except that’s not true. The reason a darker colour passes is because to a user w seeing impairments light colours (like white) bloom outward. Anyone with astigmatism knows all too well the sensation of “too much light!” interfering with the legibility of letterforms on signage. Astigmatisms are super common impairments so luckily glasses exist. This is an example of an able bodied person not understanding the entirely different perceptions impaired users can be struggling with.
A second myth listed is based on a previous post: the beloved token ui pattern previously posted. But there is an incorrect rationale here that the contrast between blue and grey is enough, because no information like error or check marks are needed since there’s no confirmation needed like in a form. Except state IS generally classified as necessary information. That’s why a checkbox passes in the first place (also why it’s more usable - let’s not even get in to the additional need to add aria labeling). Another related piece of missing contextual info is that users with cognitive and/or visual impairments often have high contrast or alternative colour modes enabled. Things like inverted colours or even black and white modes are enabled at the OS level so there’s no guarantee blue and grey are actually being presented. Semantics matter here. Without an icon or state conveyed separately the design will fail those users. There’s a veritable buffet of things accessibility standards need to be inclusive of.
Anyone who’s been through an actual a11y audit and watched the fail cases come in would generally know this stuff. So, again, get real users integrated in to these articles.
Not very useful if your org/company relies on automated testing and strict compliance with WCAG.
If your org/company is dogmatically following these myths, then this should be exceedingly useful in helping them build a better standard. But hey, complaining about free content not applying to you is easier, so that's an option too I guess.
The "myths" are the standard unfortunately.
I would recommend that designers work directly with their developers and accessibility colleagues to vet any design decisions that might go against the technical standards.
Good reading for you as well. Talks about those myths and how to solve them. Which I'm assuming is what sparked this article seeing as the stripe one was out a day before this came out on UXM.