Remove `fast-deep-equal`

[43081j]

module-replacements/docs/modules/deep-equal.md

Lines 24 to 42 in d54194b

	## `fast-deep-equal`
	[`fast-deep-equal`](https://www.npmjs.com/package/fast-deep-equal) again has the same API shape.
	Example:
	```ts
	import deepEqual from 'deep-equal' // [!code --]
	import fastDeepEqual from 'fast-deep-equal' // [!code ++]
	const a = { foo: 'bar' }
	const b = { foo: 'bar' }
	deepEqual(a, b) // true [!code --]
	fastDeepEqual(a, b) // true [!code ++]
	```
	> [!NOTE]
	> This library has a separate entrypoint for supporting Maps, Sets, etc. You can use `fast-deep-equal/es6` to support those types.

This module is actually unmaintained and has a broken ES entrypoint apparently.

I don't think we should be recommending unmaintained projects.

We could list fast-equals but its ~340KB thanks to sourcemaps and quadruple-package (cjs, umd, esm, minified)

In fact, we should probably add fast-deep-equal to the replacements itself...

[joaopedrodcf]

If it's unmainted I think it makes sense to remove it, but maybe we can also had es-tooltkit as an alternative to deep-equal as it contains isEqual and it's tree shakable and has the option if a person wants to use more utilities to not install yet another dependency

We already recommend it as an alternative to lodash and underscore

module-replacements/docs/modules/lodash-underscore.md

Line 13 in 1035393

## es-toolkit

[43081j]

we do recommend it as an alternative to lodash because lodash is already a "catch all" library. but we should really prefer avoiding catch all dependencies except for edge cases where someone genuinely uses a large amount of what they provide

so i wouldn't suggest es-toolkit as a replacement for deep-equal as it'd be a huge increase and most consumers don't need the rest of es-toolkit

granular first, catch all last - basically.