Photographers as old as I am have gone through a number of sea changes in cameras: exposure metering, autofocus, DSLR, mirrorless, and now shutterless. Curiously, Nikon has been early in all those things, though not always with their final solution (witness the Nikon 1 versus the Z System for mirrorless).
But we need to discuss what's going on with these changes, and how something fundamental is at the root of it.
Early on, automation was the primary change to cameras. Adding metering and later autofocus were both basically additions to the existing film SLR; they added automation for the user, but didn't really change the core mechanical aspects of the camera. In essence, this era was about adding electronics to what had originally been a mechanical device.
The DSLR started a process we're now deep into: silicon taking over from mechanical. This is not just a camera thing: silicon takeover has been happening across a wide range of industries and products. Why? Because producing silicon is an automated manufacturing process where costs go down with volume (the old Texas Instruments discovery). Moreover, two other trends have been going on with silicon: (1) feature size reduction that allows you to put more on a chip; and (2) the creation of additional functions on a single chip due to #1.
Initially, DSLRs didn't take full advantage of those trends, it was just more about removing mechanical film transport mechanisms and providing an all-electronic imaging system. DSLRs were still adding automation to the camera, too, which started to coalesce into a system on chip (SoC), such as Canon's DIGIC, and eventually Nikon's EXPEED, and Sony's BIONZ.
Because compact digital cameras date back to the 80's and don't have mirrors, the camera makers got a lot of experience with the imaging sensor doing "more tasks" as digital progressed. It was inevitable that as this experience advanced the camera makers would start to think about how they could apply that to interchangeable lens cameras. Mirrorless was the predictable result of that. Why? For the same reason as DSLRs replaced film SLRs: taking out a mechanical parts (mirror system) makes the cameras easier to make, and less expensive to produce, too.
Which brings us to the latest change: removing the shutter.
Nikon isn't the first one to take out the mechanical shutter, but the Z9 is the most important iteration of this change (just as the Nikon D1 was in the film-to-digital transition). Like the D1, the Z9 is a bold statement about the future. Nikon has tended to make these bold statements first in their flagship cameras, then work to move that technology down into their entire range. Canon, Fujifilm, Olympus, Panasonic, and Sony now all have to think about doing the same.
Which gets us back to why silicon is at the core of these changes now. The other camera makers have to respond because Nikon just took out parts costs, as well as assembly/maintenance/repair complexity and costs, and did so without removing function. As silicon progresses, Nikon will add function and performance.
No, I'm not going to start noshutter.com, but the change that is now about to happen to interchangeable lens cameras is not stoppable at this point.
You're probably wondering "what's next?"
Again, that's relatively predictable, though exactly how exactly it will work for cameras is still somewhat unknown; several paths exist, and it's not clear which one the Japanese will take. But look at what Apple is doing: complete SoC integration (CPU, GPU, ML, RAM, etc.). Moore's Law predicted this type of increased function, and it has clear benefits to both the producer and the user.
Back in the 70's at a tech conference I also predicted that the third dimension would have to be exploited at some point with silicon. Stacked image sensors are a basic form of that: image capture happens in one layer, data storage happens in another. Nikon themselves have shown something beyond that, though, with their 1" prototype sensor: control of the imaging layer from the second, stacked layer.
Image sensors were an easy point for exploiting the third dimension: converting a photon to an electron and doing something with that isn't intensely heat producing. It would be more problematic, for instance, to put a RAM layer below a CPU layer, for instance.
Still, I wouldn't be surprised to see another layer added, though this will have tough heat issues to deal with when you start making every layer into a "processing" layer. But I'm sure we'll see more integration into silicon, and it has to go somewhere. Some of it will go alongside the imaging area, some below.
The problem the camera companies have is one I wrote about well over a decade ago: volume. Think about this new shutterless revolution from Nikon's standpoint: they now need a DX sensor that is shutterless, as well as another FX sensor that is shutterless, at a minimum (it's possible that a Z7 III could just use the Z9 sensor, perhaps even using binned—not image binning, but the semiconductor process of selecting out "best" chips from the batch—versions that don't suss out some of the more advanced features. Still, that's a minimum of three different image sensors to get produced, for a company that's currently under the 1m annual volume level in terms of units sold. The initial cost of building that first chip is huge. You get payback by selling a ton of products with that chip.
That's one reason why Nikon tends to start their new eras at the flagship end of the product line: the higher customer cost returns pays R&D investment back reasonably fast.
_______________
Bonus: Guess what's the new "most complex" mechanical thing left in cameras? Hint: it was added recently. That's right, it's sensor-based image stabilization. Three axes of that can go away electronically (left/right, up/down, rotation ;~). Indeed, we see hints of that with Nikon's Electronic VR in video (as well as similar implementations from others). But, as fast as it was added to cameras, I have this strange feeling that it will disappear, too. Not the ability, just the mechanical parts. The tricky bit will be the tilt components of the IS mechanics. But remember, we're starting to get depth maps produced by image sensors.
