yeah sorry, I just tested¹ ...

...it's a bit counterintuitive because of the leading @ ... :-/

> @{[@$_{'VALUE','ADDRESS'}]} works great,

better don't combine both techniques, readability is more important!

you could easily do something like

my @ret=();

for ( @$table ) {
  next unless $_->{VALUE} eq $wanted;
  @ret = @$_{VALUE, ADDRESS}
}

return @ret;
[download]

To get a central exit point.

update

> if it is a little confusing to a beginner like me.

the way context is propagated into subs is always confusing, that's why this was changed in Perl6's design!

¹)

  DB<116> $_={a=>1,b=>2,c=>3,d=>0}
 => { a => 1, b => 2, c => 3, d => 0 }

  DB<117> sub tst { @$_{a,d} }

  DB<118> tst()
 => (1, 0)

  DB<119> scalar tst()
 => 0
[download]