[9.4](backport #50137) filestream: Fix shutdown logic and improve benchmark

[mergify]

Proposed commit message

Fix filestream benchmark correctness and shutdown, add high-file-count sub-benchmarks

Shutdown fix (notifyObserver):

During shutdown the watcher goroutine that drains notifyChan exits before
harvesters finish. The old blocking send in notifyObserver stalled every
closing harvester until the task group's 1-minute timeout expired. Replace the
blocking send with a select on canceler.Done() so harvesters unblock
immediately when the input is cancelled.

Benchmark fixes (correctness):

The inode-mode benchmarks were silently broken: file_identity defaulted to
fingerprint even though prospector.scanner.fingerprint.enabled was false,
so every file received the same empty-fingerprint identity and only one
harvester was started out of N files. Explicitly set file_identity.native /
file_identity.fingerprint to match the scanner mode so each file gets its own
identity and harvester.

Benchmark fixes (hangs / timeouts):

Without close.reader.on_eof: true harvesters waited for more data after EOF,
preventing the pipeline from closing until the 60-second task.Group.Stop
timeout expired. Combined with a 1-second check_interval this made multi-file
benchmarks extremely slow. Set close.reader.on_eof: true and lower
check_interval to 100 ms so harvesters close promptly.

Benchmark refactoring:

• Consolidate duplicated single-file / multi-file / inode / fingerprint
  sub-benchmarks into a table-driven loop.
• Add 1 000-file and 10 000-file fingerprint sub-benchmarks to stress per-file
  overhead (logger cloning, reader pipeline setup, fingerprint I/O).
• Replace deprecated logging setup with local loggers.
• Only buffer the event channel when events are actually collected, avoiding a
  10 000-slot channel allocation in benchmarks that discard events.

Checklist

• My code follows the style guidelines of this project
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have made corresponding change to the default configuration files
• I have added tests that prove my fix is effective or that my feature works. Where relevant, I have used the stresstest.sh script to run them under stress conditions and race detector to verify their stability.
• I have added an entry in ./changelog/fragments using the changelog tool.

How to test this PR locally

cd filebeat
go test -bench BenchmarkFilestream -benchtime 1x ./input/filestream/...
```<hr>This is an automatic backport of pull request #50137 done by [Mergify](https://mergify.com).

[github-actions]

🤖 GitHub comments

Just comment with:

• run docs-build : Re-trigger the docs validation. (use unformatted text in the comment!)

[elasticmachine]

Pinging @elastic/elastic-agent-data-plane (Team:Elastic-Agent-Data-Plane)

[mergify]

This pull request has not been merged yet. Could you please review and merge it @orestisfl? 🙏