Going to the PDFs I have made from vintage series books, here is an example of a halftone frontispiece. The image is not especially close to the gutter. Notice though that the pages are flattended on this 216-page book. On the left the tip of my letter opener can be seen. On the right is the tip of the bamboo chopstick. This is how I hold down the pages instead of the yellow finger cots. As I wrote before, they are easy for me to ignore when reading a story.
The Verne books I scanned (which are often thicker) tend to have engravings that are based on paintings so they are closer to line drawings with crosshatches than halftones.
For most of my books I prefer to retain the color of the paper as well as the system can reproduce. I use a Mac so the gamma defaults there will be different than a Windows or Linux view of the same.
I find that the color matching is not as good as a quality camera with good lighting. I have considered getting a color calibration reference to include with the photography to get a sense of adjustments that might be needed other than simply eyeballing it to what looks good.
I have used the Aura and ET24 extensively and had some good results and some that required a do over because something was not quite right.
CZUR does have a cradle on one of thier high-end models. However, I don't know if it would work as well for something like the ET24. At the neck of the unit is a set of laser line emitters that project the lines on the scanning surface or the open book. If it is on the mat, the lines are straight. When a book with curved pages is imaged, the lines show curl. This red line is interpreted by the software for the curvature flattening algorithm. Pages with red ink often confuse this so may need manual selection.
The curvature flattening only works for the open book mode of the scanner. Select the one on the right that looks like an open book. Other modes will project the 3 laser lines but don't use them.
The 3 laser lines should be all projecting on the pages of the book or you will have issues. Position the open book so they are about in the middle. If one is off the edge (straight line on the mat), this will confuse the center line calculation.
Although the software claims to let you adjust the curved laser lines, I have not had success with this. The software may hang. It is easier to reimage the page spread and get a better result the next time.
I do find that I often need to adjust the center line after imaging a spread. When I am done with a batch of pages, I return to the editor/viewer mode. Open the tool to adjust the center line and drag it to the proper place when needed. This recalcualtes both the separation of the pages but also where it will do the flattening of the images. So something that looks bad at first can often be fixed in this way.
The other thing I have to fix a lot is cropping the pages. Often the foreedge of the pages is picked up in the camera and I don't prefer to have this in my final product. There is both cropping on a per-page basis, which I normally use, and cropping for a run or selected pages. Note that the default is to clear the selection box which is almost never what you want. Instead you want to trim to the selection box. It is a bad default. A mistake cannot be undone. You can only reimage the page spread. It can be very frustrating to forget this.
Items that are very flat, such as a spiral or GBC-bound book or stack of documents can be a problem because the laser lines don't have enough data to work from to sense the center line or gutter of the book. I find that when it is confused, it picks a bad default of something like 10-15° from vertical which is almost never what you want. The magazine I showed was also flat. The solution for this was to define two page-sized rectangular selections and make sure that the publication is properly positioned before starting the imaging. It is not immediately obvious that one can have multiple page selections and I discovered it by accident when I tried to redrag a rectangle and found that I had a second one.
There is a PDF and some YouTube videos for the CZUR products. None of them are particularly advanced. The PDF is only a few dozen pages and people have asked for more. Obviously this is a Chinese product that is sold around the world. So producing a long PDF with a couple dozen translations is a bit of a chore. The marketing material from CZUR shows that nuances of English can be a challenge for them. But their English is 100x better than my French or Mandarin would be so I have to give them some credit for trying even if there are the occasional awkward phrases. We've all seen this, even among native-English speakers who are better with code or building than writing documentation. I've made a suggestion of some kind of Wiki system with trusted users having edit access to write more detailed documentation and translations of it. If CZUR does not back it, it may have to be a community project with small tips and techniques to get the most from the CZUR.
One frustration with the CZUR devices is that each one has its own application. So since I have three devices, I have three applications. As their product line grows, older devices won't get as much attention or updates to their applications. It would be wiser to have a single application that can detect the device and show the features that apply to it. Then as new features are developed, they can be available to the older devices. CZUR seems unlikely to adopt this, however. It's a bit like having a company that makes webcams and none of them are really compatible with the other. Each has their own software/driver. A lot of niche peripherals run into this problem.