Any thoughts, resources, book links that you think may be of use would be gratefully accepted!

There is a Freeware programme called 'Nuku' which unfortunately only seems to be available for Apple Macs. However it does pretty much everything you have asked in your post and maybe you have access to a Mac through work or a friend?

If so it would give you some pointers if you still want to create a Perl app with Tk (if only so that you can screengrab all the characters ;). It would also be a good visualisation tool for creating the "*fake keyboard*" layout.

Its pretty feature rich (this is from the Readme "US spelling"):

How?
Start it. Start clicking. Nuku will start in Quiz Mode the first time
you start it, the mode in which you click upon the translation of the
character shown in the upper left corner. It will measure your score,
tell you when you are wrong, and in that case, what you did wrong. Thi
+s
is the primary objective of Nuku. The other mode is Learning Mode, whe
+re
you simply click on the translated syllable you are interested in, and
Nuku will show the corresponding Japanese character.
...
Romanization systems?
There are several methods of romanizing Japanese kana, the collective
word for both hiragana and katakana. The three main ones are Hepburn,
Kunrei and Nihon-shiki. The official system is Kunrei, which does not
correspond to the actual pronounciation of the characters. However, th
+is
is the main advantage of Hepburn, it is less regular than Kunrei, but
you can also tell how to pronounce words from the romanization. The
third system, Nihon-shiki, is basically Kunrei, slightly altered to no
+t
map two kana with the same sound onto one romanization.
[download]

If you don't have access to a Mac sorry for wasting your time, but I hope this might be of some benefit if you do.

And no, I have nothing to do with the programme what-so-ever :)

I can share a program written for WinCE devices (hence strange geometry) that uses Tcl::Tk and provides English/Korenan/Japanese/Chinese dictionary searching (Japanese, Chinese taken from Monash university, thanks them!)
Also it allows searching in Unihan database.

use strict;

$::unipath='\storage card\unicode';
#my %uniexcl = ();

my $o='5669';
my $last=0;

use Tcl::Tk qw/:perlTk/;
my $mw=tkinit;
my $int=$mw->interp;
my $lfont = "-*-Arial Unicode MS--R---*-300-*-*-*-*-*-*";
my $cfont = "-*-Arial Unicode MS--R---*-200-*-*-*-*-*-*";
my $mfont = "-*-Arial Unicode MS--R---*-100-*-*-*-*-*-*";
my $tw=$mw->Text(-wrap=>'char',-font=>$cfont,-height=>4,-width=>20)->p
+ack;
$tw->tagConfigure('large',-font=>$lfont);
$tw->tagConfigure('descr',-font=>$mfont,-foreground=>'blue4');
$tw->tagConfigure('def',-font=>$mfont,-foreground=>'red4');
my $fbot=$mw->Frame()->pack;
my $ew=$fbot->Entry(-font=>$mfont,-textvariable=>\$o,-width=>10)->pack
+(-side=>'left');
my ($gdef,$ghier)=('','');
#my $e1w=$fbot->Entry(-font=>$lfont,-textvariable=>\$ghier,-width=>3)-
+>pack(-side=>'left');
my $e2w=$fbot->Entry(-font=>$mfont,-textvariable=>\$gdef)->pack(-side=
+>'left');
my $fr=$mw->Frame()->pack;
my $bw=$fr->Button(-text=>'test',
  -command=>sub{
  $tw->insert('insert',eval qq/"\\x{$o}"/);
})->pack(-side=>'left');

my $bschw=$fr->Button(-text=>'search',
  -command=>sub{
    my ($r,$what)=($tw->get('insert','insert + 1 chars'),'');
    if ($r eq '+') {
      $what=$tw->get('insert + 1 chars','insert + 5 chars');
    }
    else {
      #use Devel::Peek;
      #Dump($r);
      use Encode;
      Encode::_utf8_on($r);# will go away once Tcl bug will be fixed
      #Dump($r);
      $what=unpack("H*",encode('ucs2',$r));
    }
#print STDERR qq/=$what=\n/;
    $ghier = eval qq/"\\x{$what}"/;
    $tw->insert('end',"$ghier\n",'large');
    $tw->insert('end',$_,/kDefinition/i?'def':'descr')
      for lookup_unihan($what);
})->pack(-side=>'left');
my $bscodew=$fr->Button(-text=>'(#)',
  -command=>sub{
    my $what=$o;
    $tw->insert('end',eval qq/"\\x{$what}\n"/);
    $tw->insert('end',$_,/kDefinition/i?'def':'descr')
      for lookup_unihan($what);
})->pack(-side=>'left');
my $blookupj=$fr->Button(-text=>'l-J',
  -command=>sub{
    my $what=quotemeta $o;
    unless (defined $::jmdict) {
      open my $fhjm, "<$::unipath\\tkjmdict-0.94\\jmdict_en.bin";
      binmode $fhjm;
      local $/;
      $::jmdict=<$fhjm>;
    }
    while ($::jmdict=~/^.*?$what.*?$/img) {
      my ($t0,$t1)=($-[0],$+[0]);
      $tw->insert('end',substr($::jmdict,$t0,$t1-$t0).qq/\n/,'descr');
      $int->update;
    }
  })->pack(-side=>'left');
my $blookupch=$fr->Button(-text=>'l-CH',
  -command=>sub{
    my $what=quotemeta $o;
    unless (defined $::chdict) {
      open my $fhch, "<$::unipath\\CHIN\\cedict.utf8";
      binmode $fhch;
      local $/;
      $::chdict=<$fhch>;
    }
    while ($::chdict=~/^.*?$what.*?$/img) {
      my ($t0,$t1)=($-[0],$+[0]);
      $tw->insert('end',substr($::chdict,$t0,$t1-$t0).qq/\n/,'descr');
      $int->update;
    }
})->pack(-side=>'left');
my $blookupch=$fr->Button(-text=>'l-KR',
  -command=>sub{
    my $what=quotemeta $o;
    unless (defined $::krdict) {
      open my $fhkr, "<$::unipath\\korean\\kor2eng.ut8";
      binmode $fhkr;
      local $/;
      $::krdict=<$fhkr>;
    }
    while ($::krdict=~/^.*?$what.*?$/img) {
      my ($t0,$t1)=($-[0],$+[0]);
      $tw->insert('end',substr($::krdict,$t0,$t1-$t0).qq/\n/,'descr');
      $int->update;
    }
})->pack(-side=>'left');
my $bclr=$fr->Button(-text=>'clear',-command=>sub{
   $tw->delete('1.0','end');
 })->pack(-side=>'left');
my $bgen=$fr->Button(-text=>'gen_idx',
  -command=>sub{
    0 && do_idx();
  })->pack(-side=>'left');
my $cnt=0;

sub do_idx {
  if(!$last){
    $o='Testing... done!';
    return;
  }
  open my $fh, "<$::unipath\\unihan.txt";
  binmode $fh;
  open my $fhoidx, ">$::unipath\\unihan-idx.idx";
  binmode $fhoidx;
  my $lo='';
  my ($fpos,$fpos0)=(0,0);
  while(<$fh>) {
    $fpos=$fpos0;
    $fpos0=tell $fh;
    if (/^u\+(\w+)/i) {
      next if $1 eq $lo;
      $lo=$1;$o=$1;
      $int->update;
      printf $fhoidx "%4s%08d\n",$lo,$fpos;
#  $tw->insert('end',eval qq/"\\x{$lo}"/);
      while(<$fh>=~/^u\+$o/io){
#       print "$_";
      }
    }
    last if $last;
  #last if ++$cnt>100000;
  #  $int->update;
  }

}

$int->MainLoop;

#
# look up in unihan database.
# reads index file if necessary
sub lookup_unihan {
  my $what=shift;
  unless (defined $::idx) {
    open my $fhiidx, "<$::unipath\\unihan-idx.idx";
    binmode $fhiidx;
    local $/;
    $::idx=<$fhiidx>;
  }
  $::idx=~/^$what(.*?)$/im;
  open my $fh, "<$::unipath\\unihan.txt";
  binmode $fh;
  seek $fh, $1, 0;
  my @res;
  while (local $_ = <$fh>) {
    last unless /^u\+$what/i;
    s/\t/ /g;
    push @res, $_;
  }
  return @res;
}
[download]

Courage, the Cowardly Dog

It seems like it would be easier (for you) to store/process all the information using romaji. You could take your input this way also. You type in romaji but the japanese characters come up.

But I think I would look for an exsisting app since soon you will need to learn your vocab with kanji mixed in.

I know what you mean about wanting a customized tutorial though. On one hand, the object is to learn japanese. But unfortunately, you still have to pass the class so you have to learn THEIR vocabulary lists. ugh.

Here is a good place to learn the characters because you can learn one line at a time:
http://www.bigtrouble.com/kana