perf: add SIMD acceleration for doubleMean aggregator

[jtuglu1]

Description

We have a few use-cases of doubleMean internally. This along with min/max should finalize basic coverage of all common aggregators.

Benchmarks

Null pattern	Vector size	Scalar ns/op	SIMD ns/op	Speedup
none	128	577.647	42.667	13.54x
none	512	2252.809	198.616	11.34x
none	1024	4728.788	389.558	12.14x
none	4096	18346.941	1503.129	12.21x
sparse	128	498.490	91.306	5.46x
sparse	512	1991.397	374.181	5.32x
sparse	1024	4158.765	729.459	5.70x
sparse	4096	18250.915	2818.945	6.47x
alternating	128	257.435	91.506	2.81x
alternating	512	1105.482	378.452	2.92x
alternating	1024	2284.373	716.764	3.19x
alternating	4096	8967.622	3033.222	2.96x

Release note

Add SIMD acceleration for doubleMean aggregator

---

This PR has:

• been self-reviewed.
  • using the concurrency checklist (Remove this item if the PR doesn't have any relation to concurrency.)
• added documentation for new or modified features or behaviors.
• a release note entry in the PR description.
• added Javadocs for most classes and all non-trivial methods. Linked related entities via Javadoc links.
• added or updated version, license, or notice information in licenses.yaml
• added comments explaining the "why" and the intent of the code wherever would not be obvious for an unfamiliar reader.
• added unit tests or modified existing tests to cover new code paths, ensuring the threshold for code coverage is met.
• added integration tests.
• been tested in a test Druid cluster.

[FrankChen021]

I have reviewed the code for correctness, edge cases, concurrency, and integration risks; no issues found.

Reviewed 6 of 6 changed files.

---

This is an automated review by Codex GPT-5.6-Sol

[clintropolis]

🤘