Universal Affirmative: It may seem that "All A are B" should be rendered "Exactly ∀-0 A are B" rather than "At least ∀-0 A are B." However, "Exactly" is a compound qualifier which would mean both "At least ∀-0 A are B" and "At most ∀-0 A are B." But, as is pointed out above, "At most ∀-0 A are B" is a tautology, and so this half of "exactly" goes without saying, leaving "At least ∀-0 A are B" to carry the significant assertion. (Of course, when the exception is other than zero neither component of an "exactly claim" would be a tautology.)
Universal Negative: Also it may seem that "No A are B" should likewise be rendered "Exactly ∅+0 A are B" rather than "At most ∅+0 A are B." But as in the case above, "exactly" here again means both "At least ∅+0 A are B" and "At most ∅+0 A are B"; but since "At least ∅+0 A are B" is a tautology it leaves "At most ∅+0 A are B" as the relevant claim. (Again, with other exceptions an exact amount would not involve a tautology.)
Particular Affirmative: Since "none but 1" = "∅+1" = "plain 1" it might be less confusing initially to omit the "none but" and simply stay with tradition and say "At least 1 A are B" instead—and also simply "At most 1 A are B" for the universal negative with one exception. However, I have retained the "none but..." for symmetry's sake and also to emphasize that the numerical exceptions in either direction are deviations from polar extremes.
Particular Negative: Treating "Some A are not B" as "At most ∀-1 A are B" is clearly a departure from tradition. However, I think the origin of this peculiarity rests more with the oddness of traditional statement than with its rendition here. That is, in the traditional statements the quality of the universals is indicated by the "all" and "no" of the quantifiers while the quality of the particulars is indicated by the "are" and "are not" of the copulas; however, this practice pretty well camouflages the distinction between "Some A are not B" and its equivalent, "Some A are nonB." But by consistently using "at least" and "at most" as the qualifiers, it is clear that the proposed "At most ∀-1 A are B" is a completely different sentence from "At least ∅+1 A are nonB," and yet a moment's reflection reveals that they are equivalent nonetheless.