The oddities of the bigram "ed"

[Dunsel]

OK thanks!
This means that, while 'ed' is essentially absent on about half the pages, these pages are relatively shorter, and if one counts 'normalised pages' by splitting the longer B pages into several parts, 'ed' is essentially absent on much less than half these normalised pages.

Correct. I did not normalize the count in the one chart you liked. Now, in the third chart with the section boundaries, that was normalized by word count on the page.  And you are also correct in that the folios where ed does not occur, the word count is about 1/3 of that on the ed+ pages.  It mostly doesn't occur in herbal which is much less dense than say recipes.  But, to answer your question.

I think I got this math close. If not, tell me what you'd like to see and I'll give it a shot.

I counted the average word length of 0ed pages to 86.4 and then divided the total Voynich word count by that to get a new number of pages.  Hopefully, I averaged all that out correctly.  I also normalized the y axis on this to better show what you're looking for.   1.0 is the count of 0ed pages, so if it shows up on more, it'll have a higher count.

Here's the chart but it looks pretty much the same as the others.