Add ShowDiffraction detection denoise

[henryhng]

Summary

Detection-level denoise for ShowDiffraction.

Verification

• Checks run: PYTHONPATH=src pytest -q (1099 passed, 0 failed), npm run build, npm test (123 passed), npm run typecheck
• User workflow exercised: tutorial notebook run in JupyterLab; side-by-side Auto pipeline on low-dose real data
• Data: public (Fe3O4 tutorial SAED; pyxem Au x-grating and twinned nanowire from Zenodo 15490547) plus synthetic Poisson fixtures; 512x512 and smaller, float32
• Backend and hardware: CPU (devcontainer)
• Load/render/interaction timing: n/a (no frontend changes)
• Evidence artifacts or screenshots: interactive demos at https://henryhg.com/research/denoise and https://henryhg.com/research/denoise-model

Core checklist (every PR)

• The change includes focused tests for Python state/export behavior and
  frontend build coverage where possible; start with PYTHONPATH=src:. pytest -q
  and npm run build, or run scripts/widget_local_signoff.sh.
• Before committing, inspect git status --short and git diff --stat;
  do not commit generated HTML, docs builds, screenshots, local notebooks,
  private data, or machine-specific notes.
• Committed notebooks carry NO baked widget state (metadata.widgets) and
  pass scripts/check_notebook_sizes.py.
• Only the sections below that this PR touches are kept; the rest are
  deleted from this description.

Python API and docs — new widget, loader, or API change

• The new traits and methods have NumPy-style docs and helpful errors; no new widget is added.
• Recorded under Unreleased in CHANGELOG.md; README/API index unchanged (existing widget).
• Tutorial section added under docs/tutorials and the API page rows updated under docs/api.
• Tutorial cells let the returned widget render naturally.

---

This PR workflow follows the packaging standards for reproducible scientific
software described in scikit-package:
S. Lee, C. Myers, A. Yang, T. Zhang, Y. Xiao, and S. J. L. Billinge,
Digital Discovery (2026), DOI
10.1039/d6dd00121a.