Does the Z6 have better AF? Because the Z7 is pretty poor on AF-C - even my old D800E runs circles around it when it comes to AF-C tracking. Grabbed the camera, 70-200/4, and dog and went to the dog park to see how it would work out; for anything static the Z7 is excellent; for anything moving sideways, I'd say mostly very good; and on subjects moving towards the camera so-so. For a fast subject like a dog running towards the camera, forget it, 0% success rate on any of the AF-C modes; it just can't do it. The D800E gets about 50-70% on the same which I'd consider so-so (i.e. not an action camera if you get paid for results; I'd just get a D5/D500 for this) - the Z7 just flat out can't do it. Focus ends up about 1-2 feet behind the nose. It's not the lens (though it's not a fast focus lens) because if so the D800 couldn't do it either. So any talk of "the same AF system" etc is bull - it's not the same system. Good thing I didn't get the camera for this and I hardly ever shoot fast moving action. (If I did I'd probably get a D500.) It might be okay for sportsball - humans are a lot bigger and move slower than dogs. It might also work better with a longer lens at greater distance (less racking needed), but I think it's pretty clear it's not an action camera. It does track, it just consistently places focus too far back, indicating its control loop is too slow. (In technical terms, the error bound exceeds the range of its PID control, either through lack of measured precision or more likely inadequate sample rate.) At 5fps I ended up shooting in bursts to get sufficient VF updates, in reality 3fps is probably the highest realistic sustainable rate if you need to maintain framing. For higher, use bursts. 3fps is too slow for a dog running towards the camera. I didn't see any improvement in tracking at 3fps either, so there's no technical reason for 3 or 5 or higher, it's just about framing needs.
100% of "dog running towards camera" look like this. ISO 500, 1/800, f/4, AF-C dynamic at center splat on the dog's nose, 5fps. Results like this abound, where focus is on a rear paw or the tail, or even well behind the dog altogether.
The second problem that disqualifies it for action is I'd call the dog, it would start running towards me, I lift the camera - and the viewfinder is pitch black. From sleep mode it takes about 2 seconds to wake up, and by that time the dog is already past me, so I very quickly learned to "activate" the camera before I lift it up, by touching the shutter release (and getting about 50 random ground shots as it's a little too sensitive for this) before putting it up to my eye. The wake time is a little faster than the D800 in live view mode, so in other words ridiculously slow. This is unlike a DSLR which has VF image even if powered off, and then comes back on in a fraction of a second even if in power save. If you need to shoot opportunistically (like some PJ work) it's important to disable the viewfinder and camera power save/off modes. Just bring a pocket full off batteries; if you don't you're going to miss shots. By comparison the EVF lag is really a non-issue, it's absolutely fast enough to make no difference.
Anyway I hope someone finds this useful. Note that this is a pretty hard test that not many cameras, mirrorless or DSLR, will do well on.
Here's another example; as soon as the movement is a slower (less towards the camera) the success rate quickly climbs up to 100%. 200mm, ISO 560, 1/800s, f/4, AF-C dynamic roughly on the first dog's neck, so slightly left of center as I nudged the joystick to move it a little left for right-to-left chases, and a little to the right of center for left-to-right movement. For something like this, no problems. (Or more specifically, it's all about the photographer and light at that point.)