Well, by all means, do carry on then with your hopes of future achievement for Sony.
Meanwhile, I don't I have to "wait" for class leading excellence.
How do you like tracking wildlife with your D810?
It kind of does the job. It's not great at it.
The A9 is fantastic at it.
Now stick a high-resolution sensor into an A9 and you get a do-it-all body that outshines anything else out there - AF and cropability for action, resolution and DR for landscapes. That's what an A9r would bring. And it's a whole lot more likely to happen than a 1Dxs or D5x with the same capability.
No, I have not tried the A9 yet ... and what is a performing bird?
Falcons and parrots trained to swoop, dive and manoeuvre among and over crowds and obstacles. Also probably one of the most challenging stress tests for an AF system.
Almost certainly you will continue to live "in wanting" for many years, given your Sony fixation. After all, you are the one complaining about the behavior of the 70-200 AF performance with your Sony system as well as complaining about the mushy sharpness at f/2.8.
The problem with the 70-200 GM isn't the AF. A 70-200 GM on an A9 is just about the fastest and most accurate combination out there. Pretty much every review out there (formal or otherwise) heaps praise onto this combination for speed and accuracy.
The problem is that the edges just aren't sharp when the lens is wide-open. I don't know whether this is an optical issue or a QC/decentering issue - I suspect it's the latter, since there always seems to be at least one sharp corner or edge.
Of particular note, the reviews heaping praise on the 70-200 (especially in combination with the A9) all seem to be using it to shoot fast action. In other words, situations where edge sharpness doesn't matter at all, since the edges are going to be out-of-focus anyway. They haven't used it to shoot landscapes, or subjected it to the stress test of a brick wall or test chart wide-open. And it's true that the 70-200 GM is very sharp in the centre even wide-open, and makes a fine action lens for that reason. Just that the edge softness wide-open makes it less than stellar as a landscape lens, thus limiting its utility as a general-purpose short-to-medium telephoto lens.
Sony's AF performance issues are only when using non-native lenses via an adapter. It's still better than any other system using a non-native lens, but nowhere near as good as a native lens. In other words, it's not 'Sony's AF sucks'. It's 'AF using adapters sucks'. And it just so happens that adapters and non-native lenses are uses a lot with Sony, firstly because it's possible (you can't take a Canon 11-24 and put it on a D810, for instance) and secondly because, as a new system, Sony is still putting together its portfolio of key lenses, so users have had to adapt other lenses in the meantime.
Meanwhile, no one with a Nikon D5 + 70-200E FL ED is complaining about their autofocus speed or extreme acutense of their images. They're simply enjoying the best in the world.
Except when it comes to shooting a landscape. Then you're saddled with only 20MP and limited DR.
Switch to the D810 and your AF is no longer stellar.
Carry both and you're now carrying a pair of bricks. And you still can't make use of both the resolution and AF at the same time.
This juxtaposition between your lack-of-total enjoyment (with your system preference) and Nikonian deep enjoyment (with their system preference) will continue-on for another several years.
For a few years, maybe. Until the SLR system becomes decidedly obsolete and they ditch F-mount for something more suited to mirrorless, leaving you with a pile of lenses which are as useful on new bodies as FD lenses are today. Or if they keep F-mount but go mirrorless, leaving you with lenses that work slowly and less accurately on the new mirrorless bodies, since theit motors aren't designed for them. Or if they are unable to adapt to mirrorless and continue on as a legacy system à la Leica, becoming less and less relevant and with their performance falling further and further behind the forefront year by year. Or until Nikon collapses in a heap or are otherwise acquired by someone else, who then makes use of their patents, optical formulae and production facilities to further their own lines, leaving F-mount users orphaned.
Never invest in obsolete technology, or technology whose obsolescence can be seen coming in a few years. Would you invest in a new taxi company these days?
The only lens arena where Sony dominates is in its 85 mm recent release, which is competing with a 7-year-old Nikkor lens ... and the Sony just barely surpassed the aged Nikkor. When Nikon updates to an E lens, the Sony will be left wanting by a wide margin.
Barely surpasses it in the middle. And kicks it into the next stadium in the corners and edges.
That's one of the many problems with Lenscore. It only takes maximum sharpness into consideration. A lens can have weak edges and corners, but, so long as the middle is sharp, it will score highly in the sharpness metric.
Since you jumped from Canon to Sony, you give each more credit over Nikon than either deserves.
When I was shooting Canon (before full-frame E-mount existed), Canon lenses were the best, with the sole exception of the 14-24 (the 11-24 not having being released yet and the 16-35 Mk II having laughable edges). The 24-70 and 70-200 both beat the Nikon equivalents at the time (the 24-70 still holds its own against the current Nikon). The Canon telephotos beat the Nikon telephotos of the time (the current Nikon superteles are half a generation newer than the current Canons, with the exception of the 200-400, which, not coincidentally, is also reflected in optical performance, although none of them are slouches). The Nikon 24mm tilt-shift is a joke.
In fact, the inferiority of Nikon glass at the time was why I did not switch to the D800e when it came out in 2012, only switching to the A7r in 2014 when it allowed me to continue using the better Canon glass on a better sensor than the 5D3. The loss of AF wasn't a big deal for landscape photography, since it didn't cost me anything to keep using those lenses and I could still put them on a 1Dx to shoot action.
By the way, the best-performing Nikon bodies sensor-wise all use Sony sensors. Now Canon can also make sensors in the same performance ballpark (look at the 1Dx2 and 5D4 and extrapolate for resolution to see the likely DR of a new 5Ds2), but Nikon still can't make a sensor of its own, and the ones it designs aren't great either.
The truth is, Nikon has "enough technology" to have more class-leading cameras and lenses than Canon and Sony put together.
When Nikon decides to invest its considerable talents into a serious mirrorless system, its users will benefit from its already-superior lens portfolio.
You can't create something out of nothing. Canon and Sony both have advanced mirrorless AF systems - in Sony's case, comparable to the best SLRs - while Nikon has nothing in that field. Both Canon and Sony took years to get where they are now with mirrorless AF, and they are far bigger companies, with far more resources, than Nikon. They would literally have to stagnate for a decade if Nikon were to have a chance of catching up.
With regards to lenses, Nikon may have the upper hand with optical design (not mechanical design) at the moment, but that's only because their lenses are newer. The next generation of Canon lenses (from now and over the next 5 years or so) will almost certainly surpass the current Nikons. And the next generation of Nikons, 5-10 years from now, will almost certainly surpass that batch of Canons. It's an endless leapfrogging game, with Canon and Nikon release dates offset against each other by about half a generation.
Future Nikon mirrorless users can't benefit from the current crop of lenses. They'd need new motors to take full advantage of CDAF-based refinemeny and AI-based focus mode (eye focus, image recognition, etc.). So they'd have to buy new lenses anyway, even if the optics remained the same. Otherwise, lens performance would be like putting an A-mount lens on an E-mount camera via an adapter - not great, limited by the mechanics of the SLR lens.
Nikon hasn't produced any lemons at all, in their professional offerings, in quite a while. Mostly, they've been hitting everything out of the park (read all the top spots on SenScore/LenScore, and you will see Nikon dominating 2-1 over Canon, Leica, Sony, and Zeiss).
1. Comparing like-for-like (zoom vs zoom, prime vs prime, UWA vs UWA, etc.) notice how the top-rated lenses by sharpness (leaving aside Leica and Zeiss, who manufacture to a different level of precision - comparing Nikon and Canon here) are also the newer lenses. Not coincidentally, the newer batch of lenses are also Nikon lenses - Canon updated many of their lenses a few years back, whereas Nikon has been doing it more recently. Now strip away the newer lenses and you'll see that the next group of lenses, sharpness-wise, is that group of Canons, which are sharper than the previous group of Nikons, but not as sharp as the newest batch. Conclusion - that newer designs tend to be sharper than older ones. And, since Nikon has updated their top lenses more recently than Canon, naturally they have more in the top 10 at the moment.
2. As mentioned earlier, Lenscore is a very flawed measurement because it only considers maximal sharpness, not across-the-frame sharpness.
3. A website that rates the Canon 16-35 II as sharper than the 85/1.8, 400/5.6 and TS-E 17 has absolutely no credibility. It flies in the face of all common observation and experience. It may be accurate if they're just taking peak central sharpness into consideration (best focal length at best aperture), but that goes to show just how useless that metric is. The 16-35 II has terrible edges at all apertures. It's also very sharp in the middle. But no-one could ever argue that, overall, it's sharper than the TS-E 24L II, for example.
More useful and credible data sets provide across-the-frame data at different apertures and focal lengths. Photozone.de, for example. Even DxO (the overall marks for lenses are worse than useless and give the site a bad name, but look at the field diagrams and measured MTF charts, which show you exactly where a lens is and isn't sharp).
I guess I will conclude my debate with you by agreeing with your very last sentence.
In 10 years, Canon, Nikon, and Sony will all have awesome systems for the end-user.
The only difference is, now, only Canon and Nikon do.
In 10 years' time, Sony and Canon will likely have equal performance. Nikon will likely be making fantastic lenses for Sony cameras, as a subsidiary, or for both Sony and Canon, as a dedicated optics company. F-mount users may or may not have a legacy SLR body onto which to mount their lenses.
It's not for no reason that, even 10 years ago, when Sony had yet to make a stills camera (other than point-and-shoots), Canon saw it, rather than Nikon, as being its greatest rival in the camera business. It's come a long way in that time and has momentum behind it. Canon has the resources to match it. Nikon may be at the top of their game now, but they climbed a different peak (SLR rather than mirrorless) and probably lack the resources to go after the new objective and catch up with the two frontrunners. A Nikon-Fujifilm partnership may be able to so it, but not Nikon alone.
