Add documentation examples for ``dtype_backend="pyarrow"``

[phofl]

A while back we added an example to read_csv for dtype_backend="numpy_nullable" (see https://pandas.pydata.org/docs/user_guide/io.html#specifying-column-data-types end of the subchapter). We can add examples for dtype_backend="pyarrow" for our I/O methods as well:

• [ ] read_clipboard
• [ ] read_fwf
• [ ] read_excel
• [ ] read_html
• [ ] read_xml
• [ ] read_sql
• [ ] read_orc
• [ ] read_feather

These examples should get added to the user guide.

The following methods benefit from adding them to the docstring:

• [ ] to_numeric
• [ ] DataFrame.convert_dtypes
• [ ] Series.convert_dtypes