They have been required for a long time, but Alex Shinn says:
I definitely think everything but the one and two depth combinations should be removed from (scheme base). Their use is generally a code smell. People should use destructuring, records, or SRFI-1 first..tenth accessors.
Ray Dillinger (Bear) adds:
The historic use of these entities was as accessors for structured aggregates implemented with cons cells. In a language that directly supports records, they have a reduced mission.
On reflection, they are the Right Thing when dealing with trees built from conses. They are only problematic when used for lists.
So I'm against this migration.
The WG voted to adopt this proposal, moving c...r and c....r procedures to the (scheme cxr) library.
Arthur Gleckler is against this.
I'm on the fence.