I propose replacing the three lines of the OP (shown in sub OP) with the three lines (shown in sub TEST). The test code around them demonstrates that they produce the same result in each of the three cases of interest. The new code is about the same length and probably less efficient, but I find it much easier to understand.

use strict;
use warnings;
use Test::More;
my $y = 5;
my %Y;
my @cases = (
    [[],        'no @ARGs'],    # 1
    [[1,2,3], 'no matches'],    # undef
    [[5,2,5],  '2 matches'],    # 2
);
plan tests=>3;

foreach my $case (@cases) {
    @ARGV = @{$case->[0]};
    %Y = ();
    my $Y_OP = OP();
    @ARGV = @{$case->[0]};
    %Y = ();
    my $Y_TEST = TEST();
    is($Y_TEST, $Y_OP, $case->[1]);
}

sub OP {
    if(@ARGV) {
      foreach (@ARGV) { $Y{$y}++,next if $y eq $_;  }
    } else { $Y{$y}++ }
    return $Y{$y};
}

sub TEST{
    my $number_of_matches = grep { $_ eq $y } @ARGV; 
    $number_of_matches = undef if $number_of_matches == 0;
    $Y{$y} = (@ARGV != 0) ? $number_of_matches : 1;
    return $Y{$y};
}
[download]

OUTPUT:

1..3
ok 1 - no @ARGs
ok 2 - no matches
ok 3 - 2 matches
[download]