
That looks GREAT!!

The most important thing for me is to see that the column mode ignores column breaking figures instead of trying to change the zoom level to fit the image. From the look of it, I am guessing that 2-column mode is just setting the zoom level to 200% and then traversing the left part of the screen followed by the right half. If so, it is a simple and very useful approach to handle 2-column PDFs.
Together with the physical page/section forward button, this is a killer combo for me. No risk of accidentally changing the zoom or clicking a reference in the text that link to the bibliography on the last page (the latter was a recurring problem when trying on the kindle). I think the KOreader can do something similar, but since the Pocketbook has highlighting and note-taking in PDFs, it is the best PDF experience I have seen and read about among the 6" e-readers.
I greatly appreciate that you thouroughly went through al the modes. Based on what I saw, I will probably use column mode as the default, and when I need to view an image, I will switch to landscape mode. Actually, I might even be able to read the entire article in landscape mode depending on the text size and ability to crop margins, that is something I will need to try myself.
I guess the downsides would include a long time to open the large PDF and a little sluggish behavior when zooming in the big one, but these actions will be a small part of my overall workflow. I'm impressed it rendered that complex image in the big PDF at all, and that holding it in memory didn't seem to affect the speed of traversing the rest of the document.
This video settles it for me. I will return the Paperwhite that I ordered previously when it arrives. I ordered a Pocketbook 626 on Amazon.ca for 180 CAD (including tax and shipping), which I am now REALLY looking forward to receiving

(although the expected delivery is in about a month

).
Thank you so much again, I really appreciate that you spent the time to help me with this decision! If I find something worth mentioning in addition to your videos, I will make one myself and document the workflow I settle on for reading and annotating scientific PDFs.