docs: surface x.klickd v4.1 evidence pack in README

[Davincc77]

Summary

Polishes the public README for the now-published x.klickd v4.1 package/DOI state. Docs-only — no version bump, no tag, no GitHub Release.

• Adds a prominent x.klickd v4.1 evidence pack block near the top (above the older v3/v4 benchmark section) with the published figures:
  • 7,189 valid outputs / 7,200 expected
  • 99.85% completion across four completed bundles
  • 76.49% lower repeated input-token overhead (static x.klickd)
  • 93.34% lower with optional compressed_memory
  • automatic quality score 86.24/100 (x.klickd) vs 58.51/100 (standard AI usage without x.klickd)
  • DOI 10.5281/zenodo.20459934
• Adds npm @klickd/core and PyPI klickd latest-version badges.
• Relabels the older "The numbers" section as a historical v3.5 / v4.0 benchmark with a pointer to the current v4.1 evidence pack, to avoid reader confusion.
• Preserves claim boundaries: no universal native support, no automatic GDPR / EU AI Act compliance, no external-benchmark superiority, no industry-standard / adoption claim.

What this PR does NOT do

• Does not change package versions (already 4.1.0 on npm and PyPI).
• Does not publish, tag, or create a GitHub Release (latest Release stays v4.0.0).
• No public Chimera/codename mention (README verified clean).

Testing

• grep checks: README has zero chimera mentions and no bare forbidden-claim phrases.
• git diff --name-only confirms README.md is the only file touched.
• python3 -m pytest tests/test_v4_1_candidate_mapping.py benchmarks/v4.1/tests/ -q → 143 passed.

---

🤖 Generated by Computer