LWG 4097: views::reverse should be specialized for some view types

[jwakely]

Could LEWG (or maybe SG9?) decide if this is a change we want:
https://cplusplus.github.io/LWG/issue4097

[cor3ntin]

This was discussed here https://lists.isocpp.org/lib-ext/2024/09/27954.php

The outcome is that we need an SG-9-targeting paper that explore the design space further.
Notably

• Should repeat/empty/single/etc provide reverse iterators?
• Should we consider a type trait / cpo for "reversing is idempotent", such that we have a generic way to set / propagate that property.