Fair enough, and honestly I agree with you that supporting the n==1 case is the bare minimum of care I'd expect from an application developer.
Though if you support languages other than English even the decision to simplify pluralization logic to n==1 seems like a form of "...meh, whatever" given various pluralization rules: https://cldr.unicode.org/index/cldr-spec/plural-rules
That's certainly the approach our translation team prefers! Assembling a grammatically-correct statement from multiple discrete tokens is hard in languages with grammatical gender, complex pluralization rules, word ordering, etc. And there's not much benefit of doing it halfway right.
Though if you support languages other than English even the decision to simplify pluralization logic to n==1 seems like a form of "...meh, whatever" given various pluralization rules: https://cldr.unicode.org/index/cldr-spec/plural-rules