Fix flaky test_no_workers_timeout_queued

[crusaderky]

e.g. https://github.com/dask/distributed/actions/runs/8008396520/job/21874748036?pr=8522

>       assert s.queued or math.isinf(s.WORKER_SATURATION)
E       AssertionError: assert (<HeapSet: 0 items> or False)
E        +  where <HeapSet: 0 items> = <Scheduler 'tcp://127.0.0.1:42673', workers: 1, cores: 1, tasks: 2>.queued

note how there are only 2 tasks on the scheduler. This test relies on all submit() calls to end up in the same batched message, which is what happens most times.

[github-actions]

Unit Test Results

See test report for an extended history of previous test failures. This is useful for diagnosing flaky tests.

    26 files   -     1      26 suites   - 1   9h 35m 26s ⏱️ - 20m 7s
 3 995 tests ±    0   3 883 ✅ +    4    111 💤 + 1  1 ❌  - 4 
48 837 runs   - 1 405  46 581 ✅  - 1 362  2 255 💤  - 38  1 ❌  - 4 

For more details on these failures, see this check.

Results for commit cf77e93. ± Comparison against base commit 9bdb381.

This pull request skips 1 test.

distributed.tests.test_scheduler ‑ test_decide_worker_rootish_while_last_worker_is_retiring

♻️ This comment has been updated with latest results.

[hendrikmakait]

Thanks, @crusaderky!