the weakness of h

I'm in the midst of writing a paper on stop aspiration in Navajo, which is never regular glottal aspiration like we have in English, but instead palatalized, velarized, and/or labialized. One of the issues I'm grappling with is why the glottal fricative h is a fairly weak phoneme. I have the intuition that it is, both as a speaker of a language that has /h/, and by looking cross-linguistically at languages that have an orthographic h (and thus an historical h) but not a pronounced h, Spanish in particular.

Spanish has the letter h in its orthography, but it is never pronounced, and Spanish speakers are often unable to correctly pronounce the glottal h of English, instead substituting the velar fricative [x] found in Spanish (represented orthographically by j and sometimes g). British English has h-less dialects and h-dropping. I'm a bit puzzled by the prescriptive rule for using the article "an" before a word beginning with h, since I've been told h-dropping in British English is fairly low-class, and thus I fail to see how this pronunciation was immortalized in our rules for good writing.

These facts, combined with the relative rarity of h cross-linguistically when compared to stops or other fricatives, have given me the impression that h is indeed weakly represented somehow, most likely because it is more difficult to perceive clearly than other fricatives. However, I've so far been unable to find any good literature on the matter. I suppose I can always posit the weakness of h myself, but it's always preferable to back up one's own opinions with citations.