padam

For some reason, I just can't help seeing a parallel to this scene (around 1:40)

https://www.cameramemoryspeed.com/canon-5d-mark-iv/sd-cf-card-speed-test/ Unfortunately they haven't tested the Mark III to have a side-by-side comparison but maybe you can check some other cameras to evaluate their capabilites. But it is safe to say that it won't be that much quicker than the Mark III and since we aren't moving binned 1080p here at lower framerates there might be some issues (and the software is also different). I think it will take a while to get to a stable firmware anyway, cameras are going out quicker and quicker (Panasonic FF 4k camera next year), it is not the Mark III days anymore. For now, you can just simply donate Canon a little bit of money for the C-Log upgrade and shoot really nice (although heavily cropped) video with the 5D IV with much better ISO and usable dynamic range. Very big file sizes but that ancient codec actually works well, as you can see here.

In some way they do. Because even if it will be working, I wonder how well it can record continuously, if the bitrate is even higher than the MJPEG, which almost maxes out the data transfer rate already.

That's actually not true, maybe it is sometimes, but is it mainly dependent on the optical design itself. Oh, and the glass looks a bit fancier both on the specs sheet and in real life... I am sure Sigma could make even better f/2 ART prime lenses for instance, but people would just say 'meh, too expensive but not faster'. Generally speaking, you can design a much higher resolving lens (especially in the corners) with lower geometric distortion and far less aberrations by opting to use a slower aperture (the increase going into extreme F-stops is actually monumental). The vignetting might be more significant, but that can be corrected. This is exactly why they went with f/1.8 Z primes instead of faster. Other casing point is the Leica 50/2 APO, which still sets the benchmark (although all the rest they made is not far off in absolute terms). It will be interesting to see what the 58mm f0.95 can do. I have to say, while the 35mm f/1.8 looks very sharp, I don't find the rendering anything special in these early samples. Only seen thumbnails from the 50mm f/1.8 so far.

There is, just look at the houses on the right before punching in. If it was not rotated it would be more obvious. Also visible here:

You can find A9 video comparisons on YT, but here is one: You can see the exact same stretching. So in conclusion, this is how the computation works in the current FF cameras capable of 6k readout and it is not going to be "fixed'. I'm pretty sure that the A7SIII will have less megapixels so the internal downsampling won't be pushed to its limits. But for 4k60p they may need to upgrade the codec and processing to be able to handle that (maybe with an external recorder, but probably 8-bit 4:2:2 limited).

From some earlier comparisons, I think the A9 does the same thing, so I think it is persistent in FF 6k readout mode - and as usual the answer for the fix will be called A7SIII.

I like the argument about which camera brand is better when in fact it's been laid out that they all have their marketing strategies and other different methods of milking the customers. And they won't care what happens in the forums, they care more about global sales and maybe each other (for instance the C100II would not have received price drops if the FS5 was not released and so on...) Yes Pro features codecs etc. will eventually come down to a lower level. But much later than people expect.

This covers it well I think