[test](ivm) Remove unnecessary cloud skips from IVM suites

[yujun777]

What problem does this PR solve?

Issue Number: None

Related PR: None

Problem Summary: Several IVM regression suites unconditionally skipped cloud mode without a cloud-specific reason. Remove those guards so ordinary IVM, fallback-stream, column-drop, and row-binlog coverage runs in both deployment modes while retaining guards for transaction-path and plan-shape differences.

Related issue: #65418

Release note

None

Check List (For Author)

• Test: Diff check passed; regression suite not run
• Behavior changed: No
• Does this need documentation: No

[yujun777]

run buildall

[hello-stephen]

Thank you for your contribution to Apache Doris.
Don't know what should be done next? See How to process your PR.

Please clearly describe your PR:

1. What problem was fixed (it's best to include specific error reporting information). How it was fixed.
2. Which behaviors were modified. What was the previous behavior, what is it now, why was it modified, and what possible impacts might there be.
3. What features were added. Why was this function added?
4. Which code was refactored and why was this part of the code refactored?
5. Which functions were optimized and what is the difference before and after the optimization?

[yujun777]

/review

[github-actions]

Review status: complete for head f171f9ace68ececde8c07dde338040e572e4a459 after two convergence rounds. I am requesting changes for one test-routing issue.

Critical checkpoints:

• Goal and test proof: removing the guards correctly exposes the schema-change and MTMV row-binlog suites to the normal Cloud P0 executor, and the traced Cloud paths support their assertions. However, the two fallback-stream suites remain nonConcurrent, so the standard Cloud jobs never schedule them; two of the four hunks therefore do not provide the stated ordinary Cloud coverage. The PR also reports that the regression suites were not run, so runtime validation is still absent.
• Scope and clarity: the diff is minimal and focused; it changes only four test-mode guards and no production behavior.
• Concurrency and lifecycle: there is no production locking or lifecycle change. The fallback suites intentionally use the serial group because they toggle an unscoped all-FE debug point; running them in the normal parallel pool would interfere with unrelated IVM refreshes, which is why a Cloud serial path or test/debug-point redesign is needed.
• Compatibility, parallel paths, data, and persistence: no configuration, FE/BE protocol, storage-format, EditLog, or persistence contract changes are introduced. I traced Cloud schema change, Table Stream read states and same-transaction offset commits, MTMV compute-group selection, and companion row-binlog publication; no additional Cloud/runtime defect was substantiated.
• Tests and expected results: ordered queries and existing generated outputs are consistent, setup drops objects before use, and global debug/session state is cleaned up. No builds or test executions were attempted because the review instructions prohibit them; conclusions are static-review evidence.
• Performance and observability: no applicable production impact.
• User focus: no additional focus was provided.

The inline comment covers the identical scheduling issue in both fallback-stream suites.