There is no evidence that Tolkien consider monosyllabic nouns as a distinct class in his later writings, but such nouns are more likely to be irregular than other nouns, since they represent more basic concepts. For example, the noun má “hand” does not have a plural form *már or *mái, but uses only the partitive plural máli (VT47/6). Independent pronouns are also generally monosyllabic, so they would be declined like members of this group. Such irregularities are addressed in the entries for individual words.
Origins of monosyllabic nouns: In Common Eldarin, monosyllabic nouns had some interesting behaviors, some of which are reflected in the forms used in “modern” Quenya. In particular, consonantal nouns in Common Eldarin (CE) had a distinct subjective case, made by lengthening the vowel of the final syllable: ✶tal vs. subjective ✶tāl, ✶atar vs. subjective ✶atār (PE21/75). Thus the base form (to which inflectional suffixes were added) had a short vowel (✶tal-), subjects had a long vowel (✶tāl) and direct objects had a short vowel (✶tal).
For disyllables like ✶atār, phonetic pressures forced the last vowel to shorten after the Quenya stress shift, which caused this distinct subjective case to vanish, probably before Parmaquesta (PQ). But for monosyllabic consonantal nouns, the old subjective form was reinterpreted as the base form, and the short-vowel form reinterpreted as the stem. The end result is that a lot of Quenya monosyllabic nouns have bases with long vowels but stems with short vowels: nér (ner-) “man”, tál (tal-) “foot”.
Note that some consonantal monosyllables had long vowels in CE, and these are reflected in Quenya by consonantal nouns with long vowels in both base and stem, such as: hón “heart” < ✶khōm (Ety/KHŌ-N, PE19/102). There are also consonantal monosyllables in Quenya with short vowels in both base and stem, but these mostly had a short final vowel in CE, later lost: ✶nată > nat “thing” (VT49/30).
Another interesting feature of CE consonantal monosyllables is that sometimes used their base vowel (sundóma) for joining vowels, especially for adverbial cases (allative, ablative, locative):
Thus the fortified forms added to stems yielding monosyllabic nouns always require a dissyllabic stem with ómataima: nenesse “in (the) water” (PE21/79, from stem nēn-).
It is not clear whether this pattern survived in Quenya, though. There is a declension of tál mentioned but not fully described by Wynne, Smith, and Hostetter that implies this base-joining-vowel might have survived: “an unpublished declension of tāl, c. 1967, which gives the locative forms as talasse and talse” (VT43/16). However, since this declension remains unpublished, we don’t know whether they represent “modern” Quenya or archaic forms.
As for vocalic monosyllables, Tolkien vacillated on whether or not they were possible in CE. In some places he said they were not possible:
It is doubtful if any nouns were of TA-form. Of those that show this form in later Eldarin, some are clearly seen to have lost a consonant (h, ñ), which appears or leaves traces in derived forms, and the few others, though isolated in recorded Eldarin, probably arose in a similar manner. Thus mā “hand” is derived from a √MAH, as is seen in the Q. derivative mahta “handle”, makse < mahsi, originally meaning “handiness”. pē “mouth” is probably derived from √PEÑ (Common Eldarin: Noun Structure, 1950s, PE21/70).
In other places, he said monosyllabic vocalic nouns were rare in CE but possible:
Monosyllabics in C.E. cons. + vowel ... nouns were made only by derivative additions. Nouns without such addition were perhaps even primitively rare, and in any case few have survived or left discoverable [?traces.] These were pē “lip”, pū ?, phā “breath, puff of breath”, pī “small insect, fly”, mā “hand”, mō “person”, mbā “sheep” su, “wind” skū ?, skā ?, kā “home, house”, ñ(g)ā “terror, terrible thing”, lū “bow”, srā “body”, grā “dog”. pē, phā, pī, mbā, mō, sū prob[ably] were original nouns ... Of these pē, fā, pī, sū, *kā, *hrā, *lū survived in Q. but those marked * were [?doubled and extended] (etymological notes from 1968, VT47/35).
Assuming vocalic monosyllable were possible in CE, they seem to be an exception the rule that CE vocalic nouns could not end in ī, ū. It is unclear what their ancient plural/dual forms would have been.
Conceptual Development: In the Declension of Nouns from the early 1930s, the declensions of monosyllabic consonantal nouns were much more likely to use assimilation when adding suffixes than disyllabic or longer nouns (PE21/20-24). Tolkien also distinguished a set of (mostly vocalic) monosyllabic nouns with irregular declensions (PE21/38). These irregular nouns ended in vowels, but in Primitive Elvish had some other consonant that became final and was then lost.