[C++][Dataset] Schema evolution in Dataset scanning

[asfimport]

Overview issue for tasks related to schema evolution (will link issues)

Reporter: Joris Van den Bossche / @jorisvandenbossche

Subtasks:

• [C++][Dataset] Support null -> other type promotion in Dataset scanning

Related issues:

• [Python] Schema Evolution - Add new Field (relates to)
• [C++][Python][Parquet] We should allow for overriding to types by providing a schema (relates to)
• [C++][Dataset] Enable column renaming (in physical schema -> dataset schema) in Dataset scanning (relates to)
• [Python][C++][Parquet] Support reading Parquet files having a permutation of column order (relates to)
• [C++/Python][Dataset] Support schema evolution for integer columns (relates to)
• [C++][Dataset] Schema evolution for timestamp columns (relates to)

_{Note: This issue was originally created as ARROW-11003. Please see the migration documentation for further details.}