What's it all mean?
For the professional, computational photography is probably pretty irrelevant right now. These cameras are being positioned, so far, as pretty much the same old thing but cooler. Look, you can focus, set aperture, whatever, after you take the shot! which isn't that interesting to someone who's pretty good at doing those things properly before taking the shot. But this could change.
In camera focus stacking is will probably hit the product photography guys pretty hard, in a really good way. But that's just the same thing, done better, more conveniently. No more struggling with aperture, distance to subject, diffraction, and generally balancing all that stuff. Make this stuff here all be in focus. These guys can do it (sometimes with focus stacking!) this just makes it easier. Sometimes a lot easier.
What can you do with an even approximate 3D model of the crap in front of the camera, and a huge amount of dynamic range, and a great whack of image processing capacity?
Well, you can add and delete light sources pretty easily (for some generous value of easily). Does this mean that you don't even carry lights any more, just a couple different software-defined lighting setups in your.. whatever it is? Just shoot the CEO under any kind of generic diffuse light, against any background. The 3D means sliding the background out is trivial, throw the lights in in post. Give the PR flack a dozen different looks with a couple clicks and some fiddling with the "lights." How's that for interesting?
Now it's entirely about the moment, getting that split second in time when the CEO does that thing with his eyes and the corners of his mouth crinkle up just so and he's got the look. Nailed it. Now let's deal with the lights.
Print and 2D imaging as final output doesn't seem likely to go away any time soon, but as mobile continues its apparently inevitable march to total domination, new formats, new media, whatever that implies, are going to become A Thing.
3D information could enable, just going slightly nuts here, game-ifying photos. While we might not know what the back side of the mailbox looks like, we can move things behind it. What if ads built around these photos include elements that can interact with the objects in the photo? Or simply move through the scene naturally? Is it amusing when your logo bonks into the lamppost? Throws up mocking bunny-ears behind the CEO's head? What if I can take an ordinary (ordinary for 2019, that is) product shot, and tuck a witty easter egg into it that gives you a code for a free Starbucks drink, if you can find which object it's hiding behind?
Of course you can do this now, but you need someone to manually cut the photo up and build the semi-3D model. If I can just build the easter egg part, and slap any ordinary (2019-ordinary) photo into it, well, that's a win.
Remember when drawing lines on the football field on TV was a cool thing? That was crazy cool. Now it's completely ordinary. And expected. You can't broadcast football without 'em.