As far as I understand, strings with an encoding of US-ASCII can be used wherever an UTF-8 string can be used, so this shouldn't create any new errors.
Actually, I suspect this my have something to do with recent changes in packaging of glibc in Fedora [1]. I don't have any locales installed, except the default C.UTF-8, which works just fine:
$ LANG=C.UTF-8 ruby -e "puts ''.encoding"
UTF-8
If I install all langpacks, the command above works just fine. So it is likely third party issue. I guess that Ruby probably cannot respect just the UTF-8 part ....