When you're building a product without a focused use case, you are pulled in a ton of different directions. In AR, this means focusing on fidelity, embodied in high resolution, wide field of view visuals, powerful processing, and compelling input methods.
The real question in AR is what use cases can you hit without great fidelity? What sort of value can you unlock with a low-res postage stamp overlay and slow processor instead of full FOV? That's where the go-to market effort needs to be placed.
A similar example of this overreach was in multifunction pen devices of the 90s (General Magic, Newton, EO Personal communicator). A great counter example is Apple Watch, which didn't chase the 'smartphone on your wrist' everything device, and instead picked a few key use cases, established a beachhead, and slowly added capabilities as the technology allowed.
When a category-defining product has yet to emerge on the market, there are going to be a lot of people making predictable mistakes like this - mistiming ideas, scoping the wrong set of features, getting too excited about the wrong technologies, not leveraging their assets.
If you're a product person interested in understanding more about these factors, I wrote an essay on the subject recently: https://nickpunt.com/blog/category-defining-products/