migrate: bump ASAPPlanner pin to current main, adapt to IR restructuring

[zzylol]

What

ASAPController was renamed to ASAPPlanner, and its main branch had moved far past the rev this repo was pinned to. This re-pins asap-ir/asap-l2/asap-sketch/asap-plan/asap-frontend-promql to current ASAPPlanner main (cb70086b4c4a7ba89baf2516be81d0b192137a3a) and adapts every downstream consumer in this workspace to the IR restructuring that happened upstream in the meantime. This is a full migration, not a partial one — the workspace builds clean and the test suite passes at the same rate as before the bump.

Why this is bigger than a routine pin bump

What started as "bump the pin + do a comment sweep" turned out to require adapting to real upstream architecture changes:

• Crate consolidation: asap-ir/asap-l2/asap-sketch/asap-plan were folded into asap-types (pre_asap/post_asap modules) + a new asap-aware-mapping crate. Aliased locally as planner-types (via Cargo's package = "...") to avoid a path collision with this repo's own crates/asap_types.
• SummaryKind/SummaryParams split into per-family ExactKind/ExactParams and SketchKind/SketchParams; SummaryAgg's kind+params collapsed into one family: SummaryFamilyType enum. Threaded through sketch_algebra, optimizer, emit, physical, and data_plane's query engine.
• QueryExpr::Window removed upstream (confirmed: it never had a real producer). TimeRange{range,child} was always the actual canonical shape, so every Window match arm is rewritten against TimeRange. window_fusion.rs's recognize_windowed_sketch rewrite is a genuine bug fix, not just a rename — the old code was dead against real traffic.
• QueryExpr::LetBinding/Ref removed (workload-level CSE representation gone upstream). This repo's own dead CSE consumers (optimizer/cse.rs, intent_algebra/lower.rs) are deleted to match — both confirmed to have zero live callers (PR control_plane: adopt asap-frontend-promql for L1 parsing (Part B) #428 already switched intent_algebra to asap_frontend_promql::lower_promql directly). CommonSubexprElim (R8) now always returns None; correctness-preserving, just gives up an optimization pass.
• IR flattening: L2Expr/L3Expr/Expr<C> folded directly into QueryExpr; L3Scalar renamed ScalarValue; Predicate<C> changed from wrapping L3Expr to wrapping Box<QueryExpr<C>>.
• implement_tree_in_with(expr, &BindingScope, cost_model) collapsed to implement_tree_with(expr, cost_model) (BindingScope removed upstream).
• Aggregate.aggs renamed to Aggregate.measures.

Vendored locally (upstream deleted, this repo still genuinely needs them)

• data_plane/.../summary_exec.rs: SummaryExecutor trait + execute(), deleted upstream (ASAPPlanner#190/feat(sid): migrate PrecomputedOutput + sinks to PolicyFingerprint #197). Ported from the pre-deletion crates/sketch/src/exec.rs, adapted so find_candidates takes family: &SummaryFamilyType.
• WindowKind (crates/asap_types/src/enums.rs): deleted upstream alongside QueryExpr::Window — ASAPPlanner's scope (batch workload planning) has no use for tumbling/sliding/session flush semantics, but this repo's streaming aggregation config still does. Same shape as the old re-export, so ~145 call sites needed zero changes.
• Flat SummaryKind/SummaryParams (crates/asap_types/src/accumulator_spec.rs): this repo's AccumulatorSpec dispatch (~25 call sites in accumulator_factory) never needed the exact/sketch distinction upstream now enforces — vendored as the same 14-variant flat shape, with From impls bridging from upstream's split types where this repo's other code genuinely needs to distinguish exact vs. sketch (Materialization, BackendAggregation).

Testing

• cargo check --workspace --all-targets: clean.
• control_plane lib: 706 passed, 1 failed (pre-existing, see below).
• data_plane lib: 927 passed, 0 failed.
• asap_types lib: 42 passed, 0 failed.
• Integration suites: same pattern — all green except the same pre-existing failure and 2 tests already #[ignore]d and tracked as effective_is_cumulative misclassifies bare count()/instant Cardinality-Frequency queries as non-cumulative #431.

All 3 observed failures were verified pre-existing by checking them out against unmodified main in a separate git worktree, before any of this migration's changes:

• optimizer::rules::tests::invalid_sketch_type_override_falls_back_to_default — fails identically on unmodified main. bind_workload_typed_with_item_filter's override-re-derivation logic re-derives statistic to match the override's natural family before validity is checked, making the "invalid override falls back to default" path unreachable for the 5 canonical sketch families. Genuine pre-existing logic bug, untouched by this PR's renames (confirmed via git diff main).
• controller_plan_to_query_full_roundtrip_{cms,count_sketch}_with_heap_topk — fail identically on unmodified main; matches two adjacent #[ignore]d tests in the same file already documented as a known gap, tracked as effective_is_cumulative misclassifies bare count()/instant Cardinality-Frequency queries as non-cumulative #431.

None of these are regressions introduced by this migration.

Not in scope

A repo-wide sweep of ASAPController → ASAPPlanner prose in comments/docs was only done in files this migration otherwise touched. A full sweep across files untouched by the technical migration is left as follow-up.