feat: add outliers_percentage_target to reduce_outliers() for target-based outlier reduction [stacks on #2516]

[pidefrem]

What does this PR do?

feat: add outliers_percentage_target to reduce_outliers() for target-based outlier reduction

reduce_outliers() requires manually specifying a threshold parameter. Users must run the method repeatedly to find a threshold that achieves their desired outlier rate — a trial-and-error workflow evident in 10+ open issues about reduce_outliers behavior.

Changes:

Add an outliers_percentage_target parameter (float, 0–1): auto-searches for the threshold that achieves the target outlier percentage using binary search. Works with all 4 existing strategies (probabilities, distributions, c-tf-idf, embeddings) and requires no changes to the underlying reduction logic.

# Before: manual trial-and-error
new_topics = topic_model.reduce_outliers(docs, topics, strategy="probabilities", threshold=0.3)

# After: specify the target directly
new_topics = topic_model.reduce_outliers(docs, topics, strategy="probabilities",
                                          outliers_percentage_target=0.05)

Cannot set both threshold and outliers_percentage_target — raises ValueError. Backward compatible: defaults to None, existing behavior unchanged.

Fixes #2501

Before submitting

• This PR fixes a typo or improves the docs (if yes, ignore all other checks!).
• Did you read the contributor guideline?
• Was this discussed/approved via a Github issue? Please add a link to it if that's the case.
• Did you make sure to update the documentation with your changes (if applicable)?
• Did you write any new necessary tests?

---

⚠️ Draft — stacks on #2516. This branch includes the recalculate_representative_docs() commit from #2516 (and its upstream prerequisites) as [prereq] commits, so the diff here shows that work too. Once #2516 merges, I'll rebase onto master and the prereq commits drop out, leaving only the outliers_percentage_target change. Marking ready for review after #2516 lands.