Summary:
The catalog manager async_task_pool_ thread pool currently only allows up to the number of cores threads; this can cause problems (see linked issue) since xCluster uses that thread pool with one potentially-blocking task per DB replicated.

We fix this here by simply setting no maximum number of threads for this thread pool.

Did a little bit of tidying while at it.

Fixes #26617
Jira: DB-15994

Test Plan:
Created new integration test that previously failed and now passes:
```
ybd release --cxx-test xcluster_ddl_replication-test --gtest_filter '*.CheckpointMultipleDatabases'
```

Reviewers: xCluster, hsunder, sergei

Reviewed By: hsunder

Subscribers: sergei, ybase

Differential Revision: https://phorge.dev.yugabyte.com/D44544