If you just want to show that progress is being made, rather how far things have got, monitoring the size of the output file is easy using a thread:

Added a sleep to the display loop to prevent it from running away with the cpu. Using Win32::Sleep( 500 ); or selectundef, undef, undef, 0.5 might be better?

#! perl -slw
use strict;
use threads;
use threads::shared;

my $outfile = 'output.txt';

unlink $outfile ;

my $done:shared = 0;
async{
    system("dir /s  > $outfile");
    $done =1;
};
sleep 1 until -e $outfile;
printf STDERR "\r%d\t", -s( $outfile ) and sleep 1 until $done;
print STDERR 'Output complete';
[download]

If you can estimate the size of the output file--for example in your second command, there may be some rough correspondance between the size of the output from the first and that from the second--then you can easily convert the raw numbers into an estimated percentage. If your percentage runs over by a few percent and finishes at 105%, you'll be forgiven provided that you a) indicate that it is an estimate; b) it proves reasonably accurate.

Alternatively, add 10% to your final size estimate. When the command finishes 'early' at say ~90%, no one will complain :)

Wrapping that up into a function that takes the command, the path of the output file and an estimate of the final size is left as an exercise.