PoC: Radix 2^2 + parallel kernel

[Shnatsel]

Stacked on top of #103

Adds parallelism to the FFT kernel scaffold, which lets us parallelize the final few stages that aren't parallelized by recursion.

Recursion is still preferable because it provides natural cache locality and makes the algorithm cache-oblivious.

Performance vs main: -27% time taken on Zen4, -38% time taken on M4
measured on cargo run --profile=profiling --all-features --example benchmark -- 64 28 5

[codecov-commenter]

Codecov Report

❌ Patch coverage is 49.69040% with 325 lines in your changes missing coverage. Please review.
✅ Project coverage is 87.63%. Comparing base (3aa61ad) to head (c799608).

Files with missing lines	Patch %	Lines
src/kernels/dit.rs	40.63%	298 Missing ⚠️
src/algorithms/dit.rs	81.25%	27 Missing ⚠️

❌ Your project status has failed because the head coverage (87.63%) is below the target coverage (90.00%). You can increase the head coverage or adjust the target coverage.

Additional details and impacted files

@@             Coverage Diff             @@
##             main     #104       +/-   ##
===========================================
- Coverage   98.87%   87.63%   -11.25%     
===========================================
  Files          10       10               
  Lines        2227     2831      +604     
===========================================
+ Hits         2202     2481      +279     
- Misses         25      350      +325     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.