Be sure you aren't using photo quality from your phone as your baseline for comparison. Spending the better part of a second on post-processing and enhancements isn't an option for a low-latency video feed.

Also, given how important the cameras are for high-end smartphones, and how damn many cameras the Vision Pro has, it wouldn't surprise me if the Vision Pro's budget per camera is actually less than on a flagship smartphone.

OP here. I mean the passthrough cameras, and I mean they suck compared to the expectation of this being "AR, but rendered in VR" (I also have a 'first impressions' post that goes into more detail on this).

They're definitely not as good as the cameras on my iPhone, either.

I think the author is being far too harsh. It most certainly isn't phone quality. It most certainly is a good enough quality to do most tasks. Eating, drinking, making coffee, etc... all work perfectly fine. Fine detail is less good (reading my phone is difficult but not impossible). All things considered, they probably aren't as good as they were sold, but they are still pretty amazing.

high pixel count != good quality. With all of the smartphone cameras, the sharpness in the details is poor that the image quality is comparable to a mirrorless with half the resolution or less. And then there's low light performance and artifacts which the computational photography introduces

I'm guessing he means it falls short of "reality". The adds convey the idea as if you are purely looking at the real world around you, but the quality is very much that of a camera -- and he concedes there are some limits to what you could do with that form factor.