feat(isNotCoercible): add isNotCoercible as complement of isCoercible

[char0n]

Context

isCoercible was promoted to the public API in #3429. Many isFoo functions in ramda-adjunct have an isNotFoo complement (e.g. isSymbol/isNotSymbol, isPrimitive/isNotPrimitive).

Question: is isNotCoercible actually useful?

isNotCoercible would be complement(isCoercible), returning true only for:

• Symbol values
• Objects with no toString or valueOf in their prototype chain (e.g. Object.create(null))

That is a very narrow set. In practice, if you need to guard against a TypeError on numeric coercion you already have isCoercible for the positive path. A dedicated isNotCoercible might only add noise.

Contrast with functions that have natural complements:

• isBlank / (no isNotBlank) — similar situation
• isSentinelValue / (no isNotSentinelValue) — similar situation

Proposal

Add isNotCoercible only if there is a demonstrated use case. If the consensus is that it adds no practical value, close this issue as wontfix.

Implementation (if accepted)

// src/isNotCoercible.js
import { complement } from 'ramda';
import isCoercible from './isCoercible.js';

const isNotCoercible = complement(isCoercible);
export default isNotCoercible;