Maybe this will help you?
I don't see a requirement for recursive copying. That is possible and not that much harder.

The basic idea is to create a hash of the .ext's and then use that to either print that table or create a directory for each .ext.

Read the directory(s) all at once, create a hash table and use that to do the copies.

#!usr/bin/perl -w
use strict;

my $directory = 'C:/PerlTemp';
opendir (DIR, $directory) or die "unable to open $directory\n";

my %ext2files;  #hash of array for each extension
                #.pl => name, .txt => name etc.

foreach my $file (readdir DIR)
{
   next if ($file =~ /^\./); #skip . and .. directories
   my ($ext) = ($file =~ /\.(\w+)$/)[0]; #get "xxx/yyy.ext"
   next unless defined $ext; # no ".extension"
   #now keep track of .ext to files...
   push @{$ext2files{$ext}},$file;
}

foreach my $ext (sort keys %ext2files)
{
   # this prints the .ext ending and the number of files.
   # More than one syntax can do this, but I hope that 
   # this is understandable...
   
   print "$ext  \t", scalar @{$ext2files{$ext}}," files\n";
   
   #it is possible to use the printf() formatting strings
   #in Perl. I didn't use that here, but it is possible.
}

__END__
PL      3 files
SCP     1 files
TXT     1 files
bak     1 files
bat     4 files
bin     1 files
c   28 files
cav     1 files
cmdline     1 files
cpp     1 files
csv     4 files
dat     8 files
db      4 files
dbi     1 files
doc     4 files
exe     23 files
h   2 files
html    9 files
ico     1 files
ini     1 files
jpg     1 files
new     1 files
out     2 files
p   1 files
pdf     3 files
pl      1113 files
pll     2 files
pm      2 files
script      1 files
stackdump   3 files
temp    1 files
txt     173 files
txtC    1 files
xls     1 files
xml     6 files
zip     2 files

Process completed successfully
[download]