Simplify secondary-index store API

[forhappy]

What changed

• Removed indexed_map_production(...); every supported synchronous indexed store now uses engine.indexed_map(...).
• Removed the production/verification store-profile types, validation branch, profile-specific errors, and profile-specific adapter conformance helpers.
• Removed the derived production-profile field from indexed-map health records across Rust, UniFFI, Node, WASM, Python, Ruby, Kotlin, Swift, and Java.
• Updated all synchronous adapters, generated API inventories, design/release documentation, and the hard-cutover absence gate.

Why

Production readiness is a property of the application-selected store configuration and deployment, not a second indexed-map product. The old profile taxonomy duplicated the API and encoded incomplete deployment policy in the engine.

Impact

This is an intentional hard cutover with no compatibility alias. Callers must replace indexed_map_production(...) with indexed_map(...) and stop consuming the removed profile types and health field.

Validation

• cargo clippy --all-targets -- -D warnings
• 500 core library tests
• focused secondary-index concurrency, fault, publication, and integration suites
• SQLite, redb, PGlite, RocksDB, and SlateDB adapter suites
• UniFFI: 78 tests
• Python: 23 parity tests
• Ruby: 22 parity tests / 263 assertions
• WASM Rust compilation
• Node native Rust compilation
• Kotlin/Java Maven compilation
• binding API inventory generation/check/audit/gaps
• secondary-index hard-cutover absence gate
• formatting and diff hygiene checks