I wonder if you would be better off parsing the "CVE_data_Demo" file into a hash rather than slurping the file into an array and repeatedly greping over it. Then you would be able to do a simple hash lookup for each line of your "found_vulns_test" file. It would be useful to see a sample of the data, sanitised if necessary, to form a better judgement.

It appears that you are matching against an alternation of, for example, 'PARTIAL' or '[P]' but you should note that square brackets are regular expression metacharacters that enclose a character class; in this case a class with a single member, 'P'. They should be escaped to match literals.

...
   } elsif( /PARTIAL|\[P\]/i ) {
...
[download]

Again, I wonder if a hash might be better.

...
my %biasedConfLookup = (
   COMPLETE => 100,
   q{[C]}   => 100,
   PARTIAL  =>  50,
   q{[P]}   =>  50,
   NONE     =>   0,
   q{[I]}   =>   0,
   );
...
        $biasedConf =
           exists $biasedConfLookup{ uc $_ }
           ? $biasedConfLookup{ uc $_ }
           : 0;
...
[download]

Another point; you should always check for the success or otherwise of your open statements and it is recommended that you use the three-argument form with lexical filehandles

open my $foundVulnsFH, q{<}, q{found_vulns_test}
   or die qq{open: < found_vulns_test: $!\n};
[download]

and you should avoid using the DATA filehandle for a file you open, as filehandle of that name is opened automatically for you by the perl interpreter if your script contains data inline after __DATA__ or __END__ tags.

I hope these points are helpful.

Cheers,

JohnGG