Here's some untested skeleton code.

my @out;
open my $main_fh, '<', 'main.csv'
    or die "Can't read main.csv: $!";
while (<$main_fh>) {
    chomp;
    push @out, [ $_ ];
}
foreach my $file ( glob '200*.csv' ) {
    # you'll want to store this filename for your header
    open my $fh, '<', $file or die "Can't read '$file': $!";
    /^(\d+),(\d+)$/ && push @{ $out[$1] }, $2 for <$fh>;
}
open my $out_fh, '>', 'out.csv'
    or die "Can't write 'out.csv': $!";
print {$out_fh} $header_line_here;
print {$out_fh} join( q{,}, @{$_} ), "\n" for grep defined, @out;
[download]

Some notes:

• If your main.csv numbers are very large, you should use a hash instead of an array.
• For more general CSV-related code, use Text::CSV_XS instead.
• My use of glob might assume too much about your filenames.
• I haven't bothered to close any of my open filehandles. Don't let that stop you from doing it.
• This code assumes that every number is present in every file. Things will get very messed up for any main.csv number that appears in some files and not others.
• If you find the use of an array reference confusing, see perlreftut, perlref, and References quick reference.

I haven't tested it, but hopefully it gives you some ideas.