bugfix - preserve expected types for vocab generic calls (#735)

[dannymeijer]

Summary

This PR preserves contextual expected return types across generic call planning so expression-position vocab desugar output gets the same generic retargeting as equivalent source calls. This fixes DSL-generated calls such as orders.select(...) inferring the source row model instead of the expected output row model.

Type of change

• Bug fix
• New feature
• Refactor / maintenance
• Documentation
• CI / tooling
• RFC (adds/updates docs/RFCs/*)

Area(s)

Select the primary areas touched (used for review routing; labels are managed separately):

• Incan Language (syntax/semantics)
• Compiler (frontend/backend/codegen)
• Tooling (CLI/formatter/test runner)
• Editor integration (LSP/VS Code extension)
• Runtime / Core crates (stdlib/core/derive)
• Documentation

Key details

• User-facing behavior: Expression-position vocab blocks can desugar to generic function, helper, and method calls that use the same expected output type context as direct source.
• Internals: Added one explicit expected-return path through call checking, generic function validation, callable-value calls, stdlib module calls, and generic method validation; vocab output uses that shared planner rather than a special case.
• Risks: This touches shared generic call inference. The main risk is changing when contextual return hints bind type parameters before argument checking, so the PR includes a vocab integration regression and full release-gate verification.

Testing / verification

• make test / cargo test
• make examples (if relevant)
• incan fmt --check . (if relevant)
• Manual verification described below

Manual verification notes:

• cargo test --locked consumer_check_desugared_generic_method_call_uses_expected_return_type_issue735 --test integration_tests -- --nocapture
• cargo test --locked consumer_check_desugared_public_field_callee_call_typechecks_as_method_issue727 --test integration_tests -- --nocapture
• cargo test --locked --test cli_integration run_types_rust_callback_closures_in_every_match_arm_issue733 -- --nocapture
• cargo test --locked -p incan --lib generic_method -- --nocapture
• cargo fmt --check
• git diff --check
• make pre-commit (full gate, including smoke-test-fast and benchmark build checks)

Docs impact

• No docs changes needed
• Docs updated
• Docs follow Divio intent (tutorial/how-to/reference/explanation) where applicable

If docs updated:

• Link(s): workspaces/docs-site/docs/release_notes/0_3.md

Checklist

• I kept public docs user-focused and moved internals to contributing docs when appropriate
• I avoided duplicating canonical install/run instructions in multiple places
• I added/updated tests where it materially reduces regressions

Closes #735