[GH-3252] GeoPandas: align GeoDataFrame compatibility APIs

[jiayuasu]

Did you read the Contributor Guide?

• Yes, I have read the Contributor Rules and Contributor Development Guide.

Is this PR related to a ticket?

• Yes, and the PR name follows the format [GH-XXX] my subject.

Closes #3252.
Part of #2230.

What changes were proposed in this PR?

• Add GeoDataFrame.estimate_utm_crs by delegating to the active distributed geometry column. Bounds are computed with one distributed aggregation; only the four aggregate values are materialized on the driver.
• Align GeoDataFrame.set_crs and GeoSeries.set_crs with GeoPandas:
  • accept the epsg keyword on GeoDataFrame;
  • default allow_override to False;
  • return the mutated object for inplace=True;
  • keep direct .crs assignment as an explicit override.
• Preserve Sedona's former positional GeoDataFrame.set_crs(crs, inplace, allow_override) calls through a FutureWarning deprecation path, including validation for ambiguous positional and keyword combinations.
• Fix GeoDataFrame.from_arrow to pass to_pandas_kwargs as one named dictionary on GeoPandas 1.1+, omit it on the GeoPandas 1.0 default path, and raise a clear error when the option is requested on GeoPandas 1.0.
• Consolidate GeoSeries.total_bounds into one Spark aggregation and cache CRS resolution during UTM estimation.
• Resolve public CRS return annotations and remove obsolete set_crs overloads and comments.
• Document the allow_override=False migration under the Sedona 2.0.0 breaking changes and correct the English and Chinese GeoPandas tutorials to assign CRS only when GeoParquet metadata is absent.

How was this patch tested?

• Full GeoSeries and GeoPandas parity modules against a fresh Sedona 2.0 shaded jar and native geometry serializer: 394 passed, 1 skipped.
• Full GeoDataFrame test module: 50 passed.
• Focused compatibility matrix on GeoPandas 1.0.1: 7 passed.
• Focused compatibility matrix on GeoPandas 1.1.3: 7 passed.
• Regression coverage verifies normal, empty, and all-empty total_bounds results and exactly one Spark aggregation per call.
• All repository commit hooks passed, including Black, pyupgrade, Bandit, codespell, markdownlint, and license checks.
• git diff --check and Python 3.8 syntax validation passed.

Did this PR include necessary documentation updates?

• Yes, I used the current SNAPSHOT version, v2.0.0.
• Yes, I updated the affected API docstrings, the English and Chinese GeoPandas tutorials, and the Sedona 2.0.0 breaking-change notes.