regex para excluir somente strings em minúsculas, mas salvar strings que contenham letras maiúsculas

Simplesmente assim, usando um glob:

$ find . -type f -name '*[[:upper:]]*' 
./virt-manager-5.0.0/COPYING
./virt-manager-5.0.0/NEWS.md
./virt-manager-5.0.0/INSTALL.md
./virt-manager-5.0.0/README.md
./virt-manager-5.0.0/CONTRIBUTING.md
./virt-manager-5.0.0/DESIGN.md

O -type fobjetivo é corresponder somente aos arquivos.

Se você preferir um regex:

find .  | grep '/[^/]*[[:upper:]]'

que pode ser melhor escrito como:

find . -regextype 'egrep' -regex '.*/[^/]*[[:upper:]][^/]*$'

Observação: -type fnão é combinado com -regexlike -namepara corresponder apenas a nomes de arquivos.

Meus dois centavos

Antes de usar findo comando, com -maxdepth 1!

Aviso: o shell poderia fazer isso sozinho!

echo /tmp/MG/virt-manager-5.0.0/*[A-Z]*

/tmp/MG/virt-manager-5.0.0/CONTRIBUTING.md /tmp/MG/virt-manager-5.0.0/COPYING /t
mp/MG/virt-manager-5.0.0/DESIGN.md /tmp/MG/virt-manager-5.0.0/INSTALL.md /tmp/MG
/virt-manager-5.0.0/NEWS.md /tmp/MG/virt-manager-5.0.0/README.md /tmp/MG/virt-ma
nager-5.0.0/virtManager

ou

ls -dg /tmp/MG/virt-manager-5.0.0/*[A-Z]*

-rw-r--r-- 1 user  4401 Nov 26 20:53 /tmp/MG/virt-manager-5.0.0/CONTRIBUTING.md
-rw-r--r-- 1 user 18092 Nov 26 20:53 /tmp/MG/virt-manager-5.0.0/COPYING
-rw-r--r-- 1 user 10438 Nov 26 20:53 /tmp/MG/virt-manager-5.0.0/DESIGN.md
-rw-r--r-- 1 user  1108 Nov 26 20:53 /tmp/MG/virt-manager-5.0.0/INSTALL.md
-rw-r--r-- 1 user 30648 Nov 26 20:53 /tmp/MG/virt-manager-5.0.0/NEWS.md
-rw-r--r-- 1 user  1114 Nov 26 20:53 /tmp/MG/virt-manager-5.0.0/README.md
drwxr-xr-x 6 user  4096 Nov 26 20:53 /tmp/MG/virt-manager-5.0.0/virtManager

E você pode até mesmo iterar diretamente:

printf '%-20s %-19s %3s %4s %7s\n' File Type Lne Wrd Char;\
for file in /tmp/MG/virt-manager-5.0.0/*[A-Z]*; do
    mime=$(file -b --mime-type "$file")
    if [[ $mime == text* ]]; then
        read lines words chars < <(wc < "$file")
    else
        lines=0 words=0 chars=0
    fi
    printf '%-20s %-19s %3d %4d %7d\n' \
        "${file##*/}" "$mime" "$lines" "$words" "$chars"
done

File                 Type                Lne  Wrd    Char
CONTRIBUTING.md      text/plain          125  611    4401
COPYING              text/plain          339 2968   18092
DESIGN.md            text/plain          140 1377   10438
INSTALL.md           text/plain           42  150    1108
NEWS.md              text/plain          679 4675   30648
README.md            text/plain           30  147    1114
virtManager          inode/directory       0    0       0

Com Perle uma regex:

find . -type f | perl -F'/' -lane 'print if $F[-1] =~ /[[:upper:]]/'

Com awke uma regex:

find . -type f | awk -F'/' '$NF ~ "[[:upper:]]"'