Docs: Add ITS sensitivity checks walkthrough (follow-up #749)

[drbenvincent]

Follow-up to #749.

Objective

Provide a focused ITS example for pipeline sensitivity checks.

Scope

• Add or update ITS docs/notebook section to include sensitivity analysis
• Demonstrate at least one default and one custom check
• Explain how to interpret outcomes and next steps when checks fail

Acceptance criteria

• ITS docs include a runnable, concise sensitivity walkthrough
• Uses current pipeline API patterns
• Links back to the central guide

Parent issue: #785

[drbenvincent]

Substantively addressed by #821, which adds a runnable, focused ITS sensitivity walkthrough in docs/source/notebooks/its_post_intervention_analysis.ipynb covering:

• default check via cp.SensitivityAnalysis.default_for(cp.InterruptedTimeSeries) (PlaceboInTime)
• additional PersistenceCheck to demonstrate composing a custom suite
• pipeline API (Pipeline + EstimateEffect + SensitivityAnalysis)
• a "What to do when a check looks weak" section for failure interpretation
• link back to the central guide (sensitivity_checks.md)

After discussing scope, the consensus is not to replicate the full sensitivity walkthrough across every ITS variant (its_pymc, its_covid, its_lift_test, its_pymc_comparative, piecewise_its_pymc, its_skl) — doing so would duplicate boilerplate that #785 explicitly warned against, dilute the pedagogy of those notebooks, and create maintenance debt. Several variants (notably its_skl.ipynb) also can't run the PyMC-only checks.

Remaining follow-ups before this can be closed

• Update the "Where examples already exist" table in docs/source/notebooks/sensitivity_checks.md to point at its_post_intervention_analysis for PlaceboInTime and PersistenceCheck.
• Add a single :::{seealso} callout in its_pymc.ipynb (the canonical point-intervention ITS notebook) pointing readers to the central guide and the new walkthrough.

These can be folded into #821 or shipped as a small follow-up PR.