The author makes a good argument (and I always love reading about the nitty gritty of Apple's construction and what they not merely do, but do at incredible scale and consistency), and additionally some of the structural tradeoffs involved in ceramics may be unacceptable for general survivability in a phone-sized form factor. But even so I was surprised in his 3 option list he ignored an obvious 4th. He's correct about scaling and switchover challenges, and that at least as of right now Apple doesn't seem to have any major leaps forward in ceramic manufacturing actively deployed in a product, which in turn means iPhone-scale mass manufacturer would be incredibly expensive/difficult, even for Apple.

But that's probably part of why Apple doesn't do it in the Watch either. Instead, they have an entire lineup of Watches, and then use a very expensive, niche process only for an ultra premium skin. While I still consider it unlikely (Apple hasn't done this sort of thing in one of its main lines since the TAM in 1997), if they wanted to start the playing with an advanced material they couldn't get to scale quickly one way would be to just port the Apple Watch strategy and make it exclusive to a special high end model, ie., a ceramic "iPhone Edition" for $2-3k or whatever along with the normal color options. The 10th anniversary of the iPhone would be an occasion for that too. That strategy would have the standard advantages in terms of natural market forces: very high pricing simultaneously limits demand and ensures a large margin for R&D, handling issues, and scale out.

There are still plenty of downsides and again ZrO2 ceramics may just plain not be appropriate for the future of general Apple products regardless of manufacturing. But if they had a 10 year plan to do a major material shift in their product lines, simply reflecting reality and making the new material premium at launch then moving it down the stack seems like it'd be an option they could now pursue given that they've already gone some distance to getting the public used to the idea of differently priced "skins" on an identical internal hardware platform.