The claimed Voynich page

[eggyk]

A concave mirror projection system (or a camera lucida for that matter) could be indicated by a smoothly varying distortion across the page (or disjointed pattern of such distortions). It could also account for the combination of strong agreement in overall structure—such as in the drawing and general layout of the script—alongside the inaccuracies in finer details like individual glyphs or leaf edges. These projection systems do not produce perfectly sharp images. Even a skilled user might accurately capture the main elements of a page while struggling to resolve finer details clearly. So if they simply "did their best" and filled in glyphs and details that they couldn't make out clearly, it could easily result in the kind of mixed precision observed here.

(It could also explain the gross departure of the root system from the VMS original -- that lower section may have been entirely "off-screen". Or the person just got tired and decided to abandon the tedious equipment and free-hand his own root system.)

It is also possible that the text and the illustration were copied seperately, which may cause a different relative position across the page. The workflow would have been something like "copy text -> reposition image -> copy plant in better position" (or the reverse, with plant first and text after, or simply multiple times across the entire copying). If that is the case, extrapolating distortion between the text and the plant may not work, even if a camera lucida was used. 

Actually, some slight evidence that this is indeed the case are the words directly above the plant. They appear to me to have been squished and shortened in order to fit in the existing gaps and in order to not touch the plant. 

               

Note the y in both, one has been made smaller in order to be seperate from the plant, and the other has its tail move off to the left. dchcKhy has been shortened to something like dchthy. The word dam is also incredibly close to the leaf in the copy, surely closer than any distortion would cause? I'm open to being wrong on that though. 

       

The copier has clearly made mistakes (or choices) to put certain text or features in a different place, like the dy in the first word and different spelling across the entire text, so we will have to be as careful as possible to take that into account too.

So far, it seems you’ve been comparing the VMS folio and the copy by looking at selected heights and widths. If you’re willing to take it further, here’s an approach that might yield more insight:

• First, select several dozen anchor points from identical locations on both folios—the more points, the better. These should be clearly identifiable features that correspond between the two documents, such as specific points in the script or details in the plant drawings. Ideally, the points should be distributed across the entire page.
  
• Next, analyze the distances between pairs of these points, recording the corresponding measurements from each document. Instead of comparing just a few heights and widths, this gives you a large set of distances taken at different angles and from many locations across the page.
  
• From there, create a visual map showing the distortion vectors (both magnitude and direction) across the entire page.
  

The goal of this process would be to determine whether the distortion between the two documents forms a continuous pattern. And more specifically, a pattern that could suggest the use of an optical aid—perhaps one that was slightly tilted or periodically repositioned during the copying process.

This is a great idea if we can do it well. I can try, but that may be beyond my abilities given the inconsistencies between the two