Update Matrix4x4 paths to use the new xplat intrinsics

[tannergooding]

This just updates the existing code paths. I plan on updating a few other code paths to start using intrinsics in general in a separate PR, as well as a more general simplification to mitigate some of the inlining required for this code.

CC. @a74nh

[dotnet-policy-service]

Tagging subscribers to this area: @dotnet/area-system-numerics
See info in area-owners.md if you want to be subscribed.

[Copilot]

Pull request overview

This PR refactors Matrix4x4.Impl math kernels to use the newer cross-platform Vector128 APIs (and adds a Wasm SIMD path), replacing the prior ISA-specific and scalar fallback implementations.

Changes:

• Reworked Invert to use Vector128 xplat primitives (concat/zip/unzip/shuffle + Shuffle2) instead of the SSE-only path plus scalar fallback.
• Simplified Transpose to use Vector128.ZipLower/ZipUpper.
• Updated GetDeterminant to use a Vector128-based implementation and introduced a Shuffle2 helper with SSE/Arm64/Wasm SIMD support.

Show a summary per file

File	Description
src/libraries/System.Private.CoreLib/src/System/Numerics/Matrix4x4.Impl.cs	Migrates inversion/transpose/determinant paths to xplat Vector128 intrinsics and adds a shared Shuffle2 helper (including Wasm PackedSimd).

Copilot's findings

• Files reviewed: 1/1 changed files
• Comments generated: 3

[Copilot]

Copilot's findings

• Files reviewed: 1/1 changed files
• Comments generated: 1

[tannergooding]

@MihuBot

[Copilot]

Copilot's findings

• Files reviewed: 1/1 changed files
• Comments generated: 0 new