Benchmark Sapat Transcription Providers

[dustinvali]

Summary

• Add a long-form Daytona guide for benchmarking Sapat transcription providers inside a repeatable workspace.
• Cover OpenAI, Groq, and Azure provider setup based on Sapat current README/source behavior, including --api, --quality, --language, --prompt, --temperature, and the separate correction pass.
• Add a practical scorecard/routing-policy workflow, a new Daytona definition, a first-time author profile, and an original PNG workflow diagram.

Validation

• npx --yes markdownlint-cli guides/20260511_benchmark_sapat_transcription_providers.md definitions/20260511_definition_transcription_provider_benchmark.md authors/dustin_vali.md
• git diff --cached --check
• staged filename compliance check matching .github/workflows/content-check.yml
• Full-repo markdownlint currently reports pre-existing failures in older content files, so validation is scoped to files changed by this PR.

/claim #13