I would make it with a bit of bash pipelining:

- sort for sorting file by file
- uniq for only count one line by file
- sort again to be sure only interesting lines will be printed (with the help of uniq -c)
- perl to print interesting lines

files=(MYFILES/*.txt) 

for f in ${files[*]}
do 
sort -u $f ;
done | sort | uniq -c | 
perl -ne 'if (/[[:space:]]+'"$((${#files}+1)) "'(.*)/) {print $1."\n";
+}'
[download]

Regards,