audio: fix inter-chunk timing jitter regression in maintained audio lane

[PowerBeef]

Automated audio validation is currently non-green because the maintained jitter check is failing.

Source report: #22

Problem

The maintained audio lane is failing on inter-chunk timing jitter, which blocks a fully green source-validation pass.

Evidence

• April 18, 2026 validation report: Dev validation report: April 18, 2026 #22
• inter_chunk_timing_jitter failed with CV 1.0675 against threshold 0.5
• This was the only automated failure across the maintained source gates in that run

Current ownership

• scripts/harness_lib/audio_analysis.py
• scripts/harness_lib/audio_test_runner.py
• the maintained audio-lane artifacts and timing traces produced by python3 scripts/harness.py test --layer audio

Acceptance

python3 scripts/harness.py test --layer audio passes cleanly, with inter_chunk_timing_jitter back under threshold and optional dependency skips still treated as non-failing context.

Focused rerun

• python3 scripts/harness.py validate
• python3 scripts/harness.py test --layer audio