David Shayer writing for TidBITS:

While I don’t have any inside information about current versions of watchOS and Apple Watch hardware, there are at least four reasons to think Apple won’t support third-party watch faces any time soon, if ever.

So the article is well written, and maybe it's right, but I wanted to voice why I don't view these as, "boom, nail in the coffin, this proves why they'll never happen" arguments.

Reason #1: Battery Life

David argues that developers can't be trusted to treat battery life as well as Apple can. Fair enough, but no one is asking for them to write the whole watch face framework from scratch. It's pretty obvious that there would be something like widgets on iOS or compications on the watch itself where Apple provides some frameworks that devs can use to build watch faces, and those frameworks will do all of the low-level magic that makes watch faces be respectful to battery life. Some devs may complain that it's too restrictive, but that's the compromise.

Aside: the detail about nightly battery testing was pretty cool.

Reason #2: Buggy Code

Obviously it's important for watch faces to be reliable, but it's also important for your mapping application to get you to your destination safely. A watch that was frozen because of bad code is obviosuly bad, but again, there should be a framework that Apple controls that handles most of the hard stuff so that designers can avoid having to do all the testing that Apple does.

Hell, you used to lose all activity data if your third party workout app crashed in the middle of a run, and a couple years ago Apple improved the fitness APIs to make it so that these apps were far less likely to crash, and if they did, they could come back to life almost immediately and retain all your data.

Reasons #3 and 4: Apple’s Image and Copyright Worries

Both of these are legit, and Apple surely wants your watch face to look good. They also don't want to deal with all of the copyright issues that Google and Samsung deal with on their watch face stores.

CarPlay is a great example of how Apple could do this right with watch faces: only allow a certain number of companies to make custom watch faces. have them sign additional agreements and have them go through tougher reviews. Maybe there are literally 10 companies who are able to make watch faces. Maybe Apple can reject a watch face simply because they don't like how it looks.

Besides, there are people out there who think that the single worst thing about the Apple Watch is its watch face selection. I don't agree with them, but it's definitely the case that there are great watch face designers who don't work for Apple.

Takeaway

Maybe Apple never will allow third party watch faces. I think the biggest reason they would not is that they consider the watch faces to be a fundamental part of the product and giving that up would be like letting you install a custom launcher on iOS.

To me this argument has always feels similar to the argument that Apple can't put widgets anywhere on the iPad home screen because it's too hard or too big of a change. Apple ships incredible software, and they have amazing tools that allow developers to do spectacular things safely; an argument about how Apple can't do something falls pretty flat with me, because Apple certainly can do all of these things, it's more a matter of if they align with their strategic roadmap.