The problem has to do with how the rim is constructed. Generally, offset is calculated (near as I can tell) by rim width which is the inside dimension that the tire beads butt up against. Offset math generally moves that back and forth but it never exceeds the actual rim width.
View attachment 568883
Backspacing doesn't care about the published rim width.
This is how we all measure backspacing. Note how flat the rim lip is.
View attachment 568884
Then you have a rim like this and if they were the same width, your backspacing number would vary from the above.
Versus this-
View attachment 568885
Versus this-
View attachment 568887
Point being, if we're trying to move from a 3.5" back spaced wheel to something closer to 4" to tighten up our steering with an improved scrub radius, it is probably best to see the wheel in person and do some actual measuring.
BTW- if you swing by my house with a rim and measure it like this, just know you have invited a swift kick to the taint for being a dumbass.
View attachment 568888
The way I've always understood it is that your BS is measured striaght from the WMS to the very rearmost of the bead. like your straightedges for measuring show (excluding the limp ruler ). Offset is a measurement of how far off the centerline you are, but to convert it, you're supposed to take the nominal rim width (8" for example), add in 1" (for both beads), and then that gives you the total wheel width to divide in half and find the centerline with.
I've never known wheel construction to play a part, I've always thought you add in a half inch per bead to get the total width, since that is the "standard" (that's been the standard in all the conversion calculators at least).
I think people get misled by leaving off the bead width. For example, on a wheel known as being 15x8, they assume 4" is the centerline, and so 4" BS equal 0 offset...not correct. Have to add in the inch of bead width, centerline becomes 4.5, now a 4" BS wheel is somewhere around -12.5mm offset (offset by half an inch, negatively).
But yes, I do get your point that if the bead is designed differently, then the BS of two different wheels could be different. I would think if it was different enough due to the bead, that the offset would have been different as well, though. Since offset factors in both beads, or is supposed to.