docs(v4): SG-2 TargetTypeExpression collapse doc fold-in (swift-crab-332)

[briansrls]

Summary

• SG-2 substrate collapse (TargetTypeExpression 6-arm emitted variants → { kind: TargetTypeExprKind, node: Node } product) and TargetRecordTypeShape → TargetGenericApply fold-in for conj_form already landed on main in SG-2: TargetTypeExpressionProjection (canonical carrier + Rust row + translate consumer) #3962 (src/v4/std/target_model.dag, 06_translate.dag, rust.dag).
• This PR only aligns stale design sketches (docs/design-target-realization-canonical-home.md, docs/planning/v4-correctness-ladder-2026-05-30.md) with the ratified substrate (msg_85c90150 / M6 product carrier).

Test plan

• Doc-only; no substrate or compiler changes in this PR
• Verified TargetTypeExpression { kind, node } and conj_form: TargetGenericApply on origin/main HEAD before editing sketches

Made with Cursor

[briansrls]

Scope clarification (substrate vs docs)

The six-arm TargetTypeExprEmitted* coproduct described in the manager brief was an intermediate WIP shape on the SG-2 branch; it was collapsed to the ratified product before merge to main in #3962 (2aaec8a2a).

On current main (and this PR branch, which is main + docs only):

Requirement	Status on main
(1) TargetTypeExpression { kind: TargetTypeExprKind, node: Node }	src/v4/std/target_model.dag:138-141
(2) 06_translate dispatches on wire.kind	06_translate.dag:2207-2249 (serialize_type_expr_emitted_wire_bounded)
(3) TargetRecordTypeShape → TargetGenericApply fold-in	conj_form: TargetGenericApply at target_model.dag:146; no TargetRecordTypeShape type in tree
(4) rust.dag projection row	rust_type_expression_projection() uses TargetGenericApply; target_type_expr_emitted_wire_decode consumer in SG-2 falsification claim

rg TargetTypeExprEmitted → 0 hits repo-wide.

This PR’s diff is docs-only because git diff origin/main...HEAD -- src/v4/ is empty — re-committing identical substrate would be a no-op. SG-1 / SG-5 should base on main @ #3962+, not wait on this PR for the collapse.

Awaiting TR Mgr disposition: merge #3997 as sketch alignment, close as satisfied-by-#3962, or cite a specific main file:line still wrong.

— sent from swift-crab-332