There is a movement towards C.UTF-8 for small images (containers and
VMs).  C.UTF-8 has both size and performance improvements over the more
traditional en_US.UTF-8 locale.  (The performance improvement is
currently in upstream glibc only, but we plan to bring it to rawhide and
Fedora 35 shortly.)

However, in a world where glibc-langpack-en (or glibc-all-langpacks) is
not installed on target systems, logging in over SSH does not result in
a viable locale if the client use en_US.UTF-8 (or any other locale
except C or C.UTF-8).  This causes a severe degradation in user
experience.  It's not only that UTF-8 output does not work, there are
also frequent warning messages from various tools.  Some may even refuse
to run completely.

Most distributions send locale environment variables by default:

SendEnv LANG LC_CTYPE LC_NUMERIC LC_TIME LC_COLLATE LC_MONETARY LC_MESSAGES
SendEnv LC_PAPER LC_NAME LC_ADDRESS LC_TELEPHONE LC_MEASUREMENT
SendEnv LC_IDENTIFICATION LC_ALL LANGUAGE
SendEnv XMODIFIERS

And accept them on the server side:

AcceptEnv LANG LC_CTYPE LC_NUMERIC LC_TIME LC_COLLATE LC_MONETARY LC_MESSAGES
AcceptEnv LC_PAPER LC_NAME LC_ADDRESS LC_TELEPHONE LC_MEASUREMENT
AcceptEnv LC_IDENTIFICATION LC_ALL LANGUAGE
AcceptEnv XMODIFIERS

(Some distributions also use LC_* wildcards.)

Now that servers often use minimal installations which only support a
small set of locales (C, C.UTF-8), it makes sense to discontinue
this practice.