perl -F"\|" -anlE'%x= %x?%x:(map{$_=>$i++}@F);print "$F[ $x{NAME}]|$F
+[$x{ZIP} ]"' junk2.txt
[download]

if I wanted to change the column name(s) I want selected on the fly, could that be fit into this one-liner equation relatively easily?

If you mean "change which columns are selected", or their ordering, that's trivial:

C:\test>perl -F"\|" -anle"print join'|', @F[ 1,3, 0,2 ]" junk.csv
AGE|ZIP|NAME|CITY
23|60001|AAA|STAT
34|12345|BBB|PPOR
11|2345|CCC|TRET
[download]

If you want to select by column name, that's a bit harder to do as a one-liner:

C:\test>perl -F"\|" -anle"$n=0;if($.==1){ %names=map{$_,$n++}@F;next};
+ print join'|', @F[@names{'AGE','NAME'}]" junk.csv
23|AAA
34|BBB
11|CCC
[download]

But why would you want to do that as a one liner? (Ie. Why would you use the names rather than the column numbers directly in a one-liner?)

If you want to write a script that gets names from the command line and uses that to select columns from a file:

#! perl -slw
use strict;

my @selection = delete @ARGV[ 1 .. $#ARGV ];
my $n = 0;
my @colnames = split /\|/, scalar <>;
my %colnames = map{ $_ => $n++ } @colnames;

print join '|',@colnames[ @colnames{ @selection } ];

while( <> ) {
    print join '|', ( split '\|' )[ @colnames{ @selection } ];
}

__END__
C:\test>junk.pl junk.csv AGE NAME
AGE|NAME
23|AAA
34|BBB
11|CCC
[download]