He makes the point that noun modifiers are recognized in some varieties of traditional grammar, citing the extremely popular 19th-century grammarian Samuel Kirkham as an example.
[As an aside, if you are the sort of person who takes a depraved enjoyment in back-biting, read what another 19th-century grammarian, Gould Brown, has to say about Kirkham in the introduction to Brown's mammoth A Grammar of English Grammars.]
Liberman's point is true, of course, but my complaint was less what can be found in various traditional grammar books--and sometimes I think you can find almost any assertion if you try hard enough--than how, as a practical matter, people typically come out of K-12 classrooms equipped to view language.
In that context, Kirkham is not particularly relevant. Although there is a reprint available that appears to be targeted at the ultra-conservative home-schooling audience, Kirkham hasn't been used in classrooms in generations.
Picking a more recent book, Scope English: Writing and Language Skills, Level Z, Teachers' Edition: Scholastic (1987), here is what I find under the definition for adjectives (p. 250):
Imagine that you've just bought a few items of clothing on sale. You want to tell a friend about your purchases. Would you say something like this?
I bought a sweater, a jacket, a shirt, and a pair of pants.
If you told about your purchases in that way, you wouldn't be giving your friend a clear picture of what you bought. Here's another way to describe the things you bought:
I bought a white sweater, a jean jacket, a yellow cotton shirt, and a pair of black pants.
As you can see the words in dark print give more information about each item of clothing. Now your friend has a clearer picture of what you bought. Words used in this way are called adjectives.
An adjective is a word that modifies a noun or a pronoun.
Notice that cotton (a noun) is flagged here as an adjective. I don't know what to make of the fact that yellow is not in boldface. (Is that an error, or are the authors simplifying because of the stacked modification? Not good pedagogy either way, since it's bound to cause confusion.) The text goes on to subsume the whole class of determiners into adjectives.
Given the definition here, and the examples used, is there any doubt that students (and teachers) are being encouraged to interpret any pre-head modifier as an adjective?
Now I'm sure it's the case that many other recent textbooks give better explanations than this. But very often what students take from their education is less a matter of what appears in textbooks than how teachers present it. And based on a sample of years of students in my grammar-for-teachers course, most of them come to me (at least those who have learned anything at all about parts of speech) with more or less the understanding presented in Scope.
[Update: I just found my old copy of Warriner's, which is probably the most influential of the 20th-century schoolbook grammars (not good, mind you, just influential). Unlike the book I quoted from above, it at least mentions nouns modifying other nouns, but does so in a way that is logically incoherent. The edition I'm using is English Grammar and Composition: Revised Edition, Complete Course by John E. Warriner and Francis Griffith, Harcourt, Brace & World (1965):
Under the heading "The Adjective" (pp. 8-9):
1c. An adjective is a word used to modify a noun or a pronoun.
Nouns Used as Adjectives
Nouns are sometimes used as adjectives.
barn dance dog house house paint table tennis
When you are identifying parts of speech and you encounter a noun used as an adjective, label it an adjective.
So on the one hand, Warriner does recognize these modifiers as nouns, although he immediately tells students to call them adjectives. This directive seems to be distinguishing between the formal properties of a part of speech and its function in the sentence (which is actually sensible).
But wait! The definition for adjective above is based purely on function already. So going by Warriner's own definition, the "noun used as an adjective" qualification is wholly unnecessary. The only reason to make this statement is if there is some sense that some "words used to modify a noun or pronoun" actually fall into different categories. Which is true, but at variance with the definition.
The definition for 'adjective' here really is just a synonym for 'modifier'. And indeed, if you replace every instance of 'adjective' with 'modifier', the whole discussion makes much more sense. The only problem, of course, is that now we have a whole class of words (the real adjectives), that show formal properties which are entirely distinct from other word classes (e.g., comparative and superlative inflections), and we no longer have a label for them. It's a conceptual mess. ]