Surely you realize that there's a tremendous difference between on the one hand using solid API's and tools that are baked into the OS (and thus 'installed' on a huge number of people's devices), and on the other hand either building it yourself or relying on yet another app that needs to be installed?

Not to mention execution: my experience with Layar is that it was pretty shit for the most part. Wordlens was cool, but it's not like I can just build an app that uses Wordlens' tech, and even if I could, it would be, well, Word Lens. Not AR in the broader sense.

Everything I've seen from ARKit seems much more solid and much more practically useful than the 'floaty overlay that roughly stays in place' glorified tech demos I mostly see.