People already regularly use their phones for this kind of image capture (taking a picture as documentation, not for nostalgic memories or sharing). This seems like a positive signal. Voice doesn't get used this way, but the voice interactions are cumbersome on the phone mostly because the platform is not open enough, not because of form factor issues.

In regards to voice: there's no way to have access to the easy start mechanisms of wake words or quick button access and also control what happens before intent resolution and endpointing (i.e., deciding when the voice interaction is done). You can have your own app with its own "record" button: hard to open but you have control of what happens. Or you use the assistant infrastructure and have to compete with every other Apple/Google product goal and parsing approach, and at best you have a chance to do further recording only after the initial intent has fully resolved, the mic has closed, and you can reopen it.

So yes a phone can do all that, and it ALSO would be awesome, but just like with these glasses you can't ACTUALLY implement this.