Add default tiling matching index range to MDRange loop pattern.

[pgrete]

PR Summary

Addresses performance issue observed by @jdolence when using par_reduce.
The latter uses an MDRangePolicy, which, by default, does not come with tiling information.
That results in poor performance on devices as observed before, see, e.g., https://gitlab.com/pgrete/kathena/-/merge_requests/58
The current default may not be perfect for all situations and MeshBlock sizes but it's for sure a much better default choice than "no tiling".
Partially addresses #466

PR Checklist

• Code passes cpplint
• New features are documented.
• Adds a test for any bugs fixed. Adds tests for new features.
• Code is formatted
• Changes are summarized in CHANGELOG.md
• (@lanl.gov employees) Update copyright on changed files

[pgrete]

@par-hermes format