sql: Fix autorouting for expensive functions

[def-]

Only top-level expensive functions were considered in the auto-routing

Broken in #20994

One line fix, the rest is adding tests that actually made me find this. Without the production code change the tests fail like this:

    EXPLAIN OPTIMIZED PLAN WITH(humanized expressions) AS VERBOSE TEXT FOR
    (SELECT 1) UNION ALL (SELECT 1 + 2)
    OutputFailure:test/sqllogictest/explain/auto_route_expensive.slt:41
            expected: Values(["Explained Query (fast path):\n  Constant\n    - (1)\n    - (3)\n\nTarget cluster: quickstart\n"])
            actually: Values(["Explained Query (fast path):\n  Constant\n    - (1)\n    - (3)\n\nTarget cluster: mz_catalog_server\n"])
            actual raw: [Row { columns: [Column { name: "Optimized Plan", table_oid: None, column_id: None, type: Text }] }]
    ----
    EXPLAIN OPTIMIZED PLAN WITH(humanized expressions) AS VERBOSE TEXT FOR
    SELECT 1 WHERE EXISTS (SELECT 1 + 2)
    OutputFailure:test/sqllogictest/explain/auto_route_expensive.slt:54
            expected: Values(["Explained Query (fast path):\n  Constant\n    - (1)\n\nTarget cluster: quickstart\n"])
            actually: Values(["Explained Query (fast path):\n  Constant\n    - (1)\n\nTarget cluster: mz_catalog_server\n"])
            actual raw: [Row { columns: [Column { name: "Optimized Plan", table_oid: None, column_id: None, type: Text }] }]
    ----
    EXPLAIN OPTIMIZED PLAN WITH(humanized expressions) AS VERBOSE TEXT FOR
    SELECT 1 WHERE (SELECT 1 + 2) > 0
    OutputFailure:test/sqllogictest/explain/auto_route_expensive.slt:66
            expected: Values(["Explained Query (fast path):\n  Constant\n    - (1)\n\nTarget cluster: quickstart\n"])
            actually: Values(["Explained Query (fast path):\n  Constant\n    - (1)\n\nTarget cluster: mz_catalog_server\n"])
            actual raw: [Row { columns: [Column { name: "Optimized Plan", table_oid: None, column_id: None, type: Text }] }]
    ----
    FAIL: output-failure=3 success=3 total=6

[github-actions]

Thanks for opening this PR! Here are a few tips to help make the review process smooth for everyone.

PR title guidelines

• Use imperative mood: "Fix X" not "Fixed X" or "Fixes X"
• Be specific: "Fix panic in catalog sync when controller restarts" not "Fix bug" or "Update catalog code"
• Prefix with area if helpful: compute: , storage: , adapter: , sql:

Pre-merge checklist

• The PR title is descriptive and will make sense in the git log.
• This PR has adequate test coverage / QA involvement has been duly considered. (trigger-ci for additional test/nightly runs)
• If this PR includes major user-facing behavior changes, I have pinged the relevant PM to schedule a changelog post.
• This PR has an associated up-to-date design doc, is a design doc (template), or is sufficiently small to not require a design.
• If this PR evolves an existing $T ⇔ Proto$T mapping (possibly in a backwards-incompatible way), then it is tagged with a T-proto label.
• If this PR will require changes to cloud orchestration or tests, there is a companion cloud PR to account for those changes that is tagged with the release-blocker label (example).

[ggevay]

Yes, thank you, this is definitely fixing a bug!

(I hope fixing this won't break somebody's dbt workflows. The PR will make some queries not run on mz_catalog_server, but on the active cluster, and 6a93648's commit msg says that the auto-routing of simple queries was introduced to "help some dbt use cases". I hope those dbt queries will still be covered. Unfortunately that commit doesn't seem to have added tests for the specific dbt queries. Tbh, this whole auto-routing feels quite messy.)