Likewise, with windows at about R-3, I really wonder how much it helps to go from R17 to R22 for example.
It's helpful to remember that R-values are really reciprocals of what you
probably want to know: they're degrees square-feet per btu/hour (also
known as square-feet degree hours per btu). The reciprocal of R value,
known as U value, is more useful, I think. Because you can have your
entire house insulated to R-100, except that 1 square foot is insulated
to R-epsilon (where 'epsilon' is an incredibly tiny number), and your
house is very poorly insulated. But if you look at U-values, most of
your envelope is U=0.01, but 1 sq-ft is at U-gazillion; so when you add
it all up, the overall btu/hour per degree is still pretty big, and thus the
overall R-value is pretty low. Not exactly realistic, but it makes
the point (I hope).
Similarly, we measure fuel economy in MPG. Apparently they use a
a reciprocal measurement, something like liters per 100 kilometers,
in Canada. That's more helpful. For example, if you ask the
average non-math-geek person whether it saves more gas to
increase somebody's gas mileage from 10 to 15 MPG, or from 20 to
40 MPG, they'll probably answer the latter. (If we used "gallons per
hundred miles" as the fuel-economy measure, you'd more easily see
that the first person goes from 10 to 6-2/3, whereas the second goes
from 5 to 2-1/2 - so the first saves more gas).
Of course, it's also true that R-values are more intuitive when you're
looking at "series" thermal "circuits". So you can add the R-value of
that outer foam layer to the R-values of the stud spaces + batts. With
U-values, you have to add them as reciprocals - like you do resistors
in an electrical circuit, and like you have to do R-values for elements
that are in parallel (like my R-100 house with a small area with
extremely small R-value).
Hope this all makes sense and is of some interest and helpfulness ...