What am I not getting?

Now I looked, you also need to read https://metacpan.org/module/HTML::TreeBuilder#parse_file because treebuilder is interpreting those UTF-8-encoded-bytes as latin-1

This works

#!/usr/bin/perl --
use strict;
use warnings;
use autodie;

use WWW::Mechanize 1.72;

#~ use HTML::TreeBuilder::XPath;
#~ use HTML::TreeBuilder::LibXML;


Main( @ARGV );
exit( 0 );

sub xtree {
    local $@;
    if( eval { require HTML::TreeBuilder::LibXML; } ){
        return HTML::TreeBuilder::LibXML->new;
    }
    if( eval { require HTML::TreeBuilder::XPath; } ){
        return HTML::TreeBuilder::XPath->new;
    }
    die "$@
you need to install   use HTML::TreeBuilder::XPath
or    use HTML::TreeBuilder::LibXML\n\n";
}

sub Main {
    my $url = shift or die "\n\nUsage: $0 http....\n\n";
    binmode STDOUT, ':encoding(UTF-8)'; ## grr
    
    my $mech = WWW::Mechanize->new( autocheck => 1 );
    my $tree = xtree();

    $mech->get( $url );
    $tree->parse( $mech->content );
    
    for my $node ( $tree->findnodes( q{ //span[ @itemprop="reviewBody"
+ ] } ) ){
        print $node->as_HTML, "\n\n";
        print $node->as_text, "\n\n";
        print $node->as_HTML( q{<&>} ), "\n\n";
    }
}
__END__
[download]