OK, so I've re-sized my 4 pages so that the x-height is 7 mm in all of them.

I've noticed that the text block size varies! It can be seen by clicking on the scans shown in Script Analysis - Ascender and Descender Height.

I've taken the first image - the page shown in Backhouse's Illuminated Page, which is also the page shown in the Historical Source book for Scribes, as a basis.

I've then marked on each of the other pages the differences in the height and width of the text block. There are differences varying between 8 mm and 19 mm on the various pages.

I can explain why the other 3 pages are shorter. The median (most common) ascender/descender height for the 2 British Library pages is 2 mm, not 3 mm. The median ascender/descender for the Codices Illustres page is 3 mm, but the measurements are weighted differently - there are a lot of smaller ascender/descender heights on that page.
So having a lot of smaller ascender/descender heights on a page ends up in a shorter page.

Why the lines are shorter on the other 3 pages - I just don't know. The writing on the Backhouse page looks a little bigger. It has the most flourished version of the script - I've mentioned before that it looks like the pages were done by different scribes, most noticeable with the Backhouse page - which is why it might be spaced out a tiny tiny bit more. I haven't had any success in finding a difference through measurement - the measurements are just too small.

I'm not going to worry about it too much. I've measured the x-height, and the ascender/descender height and there is so much variability on one page, never mind between the pages. I just need to settle on a 'standard' that I can work from.