fix(color): recover numeric palette slots lost to transparency

[phil-r]

Description:

Fix numeric color requests losing a visible color because hidden RGB in transparent pixels consumes a palette slot. Both color profiles retry only when an occupied cluster has no visible pixels before component cleanup. Recovery preserves the numeric count and initialization method, cleans only the final pass, and leaves input arrays unchanged. Cleanup-only losses do not trigger recovery; auto and binary behavior stay unchanged.

This PR includes:

• Regression tests for hidden RGB, numeric counts, cleanup-only losses, unused centroids, empty images and repeated solves.
• Updated numeric-count documentation and HTTP profile checks, retaining the independent original-pipeline SVG hash.

How to run:

docker compose --profile testing run --build --rm test \
  python -m pytest vectorizing/tests/test_color_numeric.py \
  vectorizing/tests/test_color_configurations.py

Validation: 827 offline tests and all pre-commit hooks passed. test_integration.py (HTTP/S3/gallery integration) was excluded. Unaffected SVG, RGBA, palette, dimension and bounds captures matched exactly. Visual evidence remains local; no artwork or gallery baselines were uploaded or replaced.

Local benchmarks on the reported PNG (M4 Pro, ARM64 Docker/Python 3.11, single-threaded, 5 fresh-process samples plus warmup):

Profile	Original, 2 colors	Fixed, 3 colors
current	213 ms	352 ms
experimental	169 ms	271 ms

Recovery adds a bounded extra clustering pass. No consistent slowdown was measured on unaffected inputs. Timing excludes decoding, HTTP/S3 and rasterization; production concurrency was not tested.

Checklist (based on PR guidelines):

• I have self-reviewed this PR, according to these points
• This PR falls into maximum three of these categories
• I feel comfortable taking responsibility for merging this code to production
• The code is high quality, well-tested, follows styles guides (Frontend), clean, and well documented