GH-50496: [R] Polish NEWS.md for 25.0.0

[thisisnic]

Rationale for this change

Polish NEWS.md for the 25.0.0 CRAN release.

What changes are included in this PR?

Organized and expanded NEWS entries following tidyverse style.

Are these changes tested?

No code changes, NEWS only.

Are there any user-facing changes?

No.

• GitHub Issue: [R] Polish NEWS.md for 25.0.0 #50496

[github-actions]

⚠️ GitHub issue #50496 has been automatically assigned in GitHub to PR creator.

[Copilot]

Pull request overview

Polishes the R package NEWS.md entry for the Arrow 25.0.0 CRAN release by reorganizing the release notes into tidyverse-style sections for improved readability and consistency.

Changes:

• Adds structured subsections for 25.0.0 (Breaking changes, New features, Minor improvements and fixes, Installation).
• Reflows and groups existing bullets under the new headings.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 2 comments.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 1 comment.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

[jonkeane]

Thanks for this, oops that we didn't merge before the CRAN release!

[conbench-apache-arrow]

After merging your PR, Conbench analyzed the 4 benchmarking runs that have been run so far on merge-commit 32562bd.

There was 1 benchmark result indicating a performance regression:

• Commit Run on test-mac-arm at 2026-07-20 02:04:00Z
  • tpch (R) with engine=arrow, format=native, language=R, memory_map=False, query_id=TPCH-07, scale_factor=1

The full Conbench report has more details. It also includes information about 78 possible false positives for unstable benchmarks that are known to sometimes produce them.