In this case I would say that regardless of intent (which may still be important) we should go for the better image format if we have the option. Like, to take the historical example mentioned in the NYT article linked above, this is obviously bad:
> Concordia University professor Lorna Roth’s research has shown that it took complaints from corporate furniture and chocolate manufacturers in the 1960s and 1970s for Kodak to start to fix color photography’s bias.
.. but without excusing any of that, black people were still better off with the newer films.
Similarly, regardless of the intent of the JXL design, if it is the more "inclusive" tech then that speaks in favor of it.
Also, the difference in the example images requires you to look actively look for JPG artifacts. That's something that usually only codec developers and people who work with images for their day job do. Almost everyone else subconsciously does the opposite: they learn how to "look past" the compression artifacts. But speaking as someone who did spend a lot of time trying to optimize JPG sizes for a short while in his life there is definitely a difference.
The difference is hard to see when compressing an image once, but keep in mind that we live in a world where images go through lossy recompression dozens of times, causing the "needs more JPEG" meme[0][1] to live on even when it should be a thing of the past. If your image is, say, 5% more visually lossy for black people that adds up.
> Concordia University professor Lorna Roth’s research has shown that it took complaints from corporate furniture and chocolate manufacturers in the 1960s and 1970s for Kodak to start to fix color photography’s bias.
.. but without excusing any of that, black people were still better off with the newer films.
Similarly, regardless of the intent of the JXL design, if it is the more "inclusive" tech then that speaks in favor of it.
Also, the difference in the example images requires you to look actively look for JPG artifacts. That's something that usually only codec developers and people who work with images for their day job do. Almost everyone else subconsciously does the opposite: they learn how to "look past" the compression artifacts. But speaking as someone who did spend a lot of time trying to optimize JPG sizes for a short while in his life there is definitely a difference.
The difference is hard to see when compressing an image once, but keep in mind that we live in a world where images go through lossy recompression dozens of times, causing the "needs more JPEG" meme[0][1] to live on even when it should be a thing of the past. If your image is, say, 5% more visually lossy for black people that adds up.
[0] https://knowyourmeme.com/memes/needs-more-jpeg
[1] http://needsmorejpeg.com/