You're on the right track with your question. The mostly duplicate code that handles the two tables should probably be put into a subroutine which your main script can then call and pass the header array references to as an argument:

sub get_table {
    my $headers = shift;
    $table_extract = HTML::TableExtract->new(headers => $headers);
    $table_output = Text::Table->new();
     
    $table_extract->parse_file($html);
    ($table) = $table_extract->tables;
     
    foreach my $row ($table->rows) {
        $table_output->load($row);
        print "   ", join(',',grep defined, @$row),"\n";
        print $ofh "   ", join(',',grep defined, @$row), "\n";
    }
}

# calling the sub for each table
get_table( [ 'Status', 'Results', 'Schedule Name' ] );
get_table( [ 'Status', 'Results', 'Node Name' ] );
[download]