In the following code, I've cut down the input data to the minimum required to demonstrate the technique that I've used.

#!/usr/bin/env perl

use strict;
use warnings;

use Data::Dump;

my $data = [
    { status => 'Main',  City => 'NY', zCode => '0002', ID => '2222' }
+,
    { status => 'Extra', City => 'BO', zCode => '0007', ID => '2222' }
+,
    { status => 'Cabin', City => 'NE', zCode => '4562', ID => '2222' }
+,
    { status => 'Main',  City => 'AA', zCode => 'TTTT', ID => '8888' }
+,
    { status => 'Test',  City => 'BB', zCode => 'WWWW', ID => '8888' }
+,
    { status => 'Cabin', City => 'CA', zCode => '2334', ID => '8888' }
+,
];

my %new = (City => 'new_City', zCode => 'new_z_code');
my %extra;
$extra{$_->{ID}} = $_ for grep +($_->{status} eq 'Extra'), @$data;

for my $row (@$data) {
    next unless $row->{status} eq 'Main';
    %$row = (%$row,
        map +($new{$_} =>
            exists $extra{$row->{ID}} ? $extra{$row->{ID}}{$_} : ''
        ), keys %new
    );
}

dd $data;
[download]

Output:

[
  {
    City => "NY",
    ID => 2222,
    new_City => "BO",
    new_z_code => "0007",
    status => "Main",
    zCode => "0002",
  },
  { City => "BO", ID => 2222, status => "Extra", zCode => "0007" },
  { City => "NE", ID => 2222, status => "Cabin", zCode => 4562 },
  {
    City => "AA",
    ID => 8888,
    new_City => "",
    new_z_code => "",
    status => "Main",
    zCode => "TTTT",
  },
  { City => "BB", ID => 8888, status => "Test", zCode => "WWWW" },
  { City => "CA", ID => 8888, status => "Cabin", zCode => 2334 },
]
[download]

You should only need to add two key-value pairs to %new for a full solution.

Consider simply adding 'new_' to the existing keys for a more standard naming convention. This would make comparing keys easier; if not a requirement now, it could be at a later date. It would also allow you to simplify my code, as follows:

my @new = qw{City zCode ...};
...
        map +("new_$_" =>
            exists ...
        ), @new
[download]