Yet another solution:

use XML::Rules;

my $parser = XML::Rules->new(
  rules => [
    item => sub { '<li>' . $_[1]->{_content} . '</li>'},
    list => sub {
      for ($_[1]->{_content}) {
        s/^\s+//;
        s/\s+$//;
        s{</li>\s+<li>}{</li><li>}g;
      }
      "\nTX <ol>" . $_[1]->{_content} . "</ol>\n"
    },
    para => sub { 'TX ' . $_[1]->{_content}},
    head => sub { 'TX <h3>' . $_[1]->{_content} . "</h3>"},
    'doc.body' => 'content',
  ]
);

use Data::Dumper;

my $result = $parser->parse($XML)->{'doc.body'};
for ($result) {
    s/^\s+//;
    s/\s+$//;
}

print $result;
__END__
C:\temp>c:\temp\reformatXML.pl
TX <h3>Introduction</h3>
TX para 1
TX para 2

TX <h3>Individual</h3>
TX para 1
TX para 2:

TX <ol><li>item 1</li><li>2</li></ol>
[download]

Getting rid of the excess whitespace complicates the code somewhat.