I hacked together a first, unfinished, largely untested prototype converter. This goes into the base directory of https://sourceforge.net/p/pdl/pdl-book/ci/master/tree/ as make_epub.pl:

#!/usr/bin/env perl

use v5.36;

use strict;
use warnings;

use Carp;
use App::Pod2Epub;
#use Archive::Zip;
use GD;
use MIME::Base64;
use English;

my $parser = App::Pod2Epub->new();

my @full;
my @subfiles;

{ # Read and parse TOC pod
    my @lines = _readFile('PDL/Book.pod');
    
    my $istoc = 0;
    foreach my $line (@lines) {
        if($line =~ /^\=over/) {
            $istoc = 1;
            #push @full, $line;
            next;
        } elsif($line =~ /^\=back/) {
            $istoc = 0;
            #push @full, $line;
            next;
        } elsif($istoc && $line =~ /^\=item/) {
            my $subfile = '';
            if($line =~ /\<(.+?)\|/) {
                $subfile = $1;
                $line =~ s/.*\>\ //;
                $subfile =~ s/\:\:/\//g;
                $subfile .= '.pod';
                if(!-f $subfile) {
                    warn("$subfile does not exist");
                } else {
                    push @subfiles, $subfile;
                }
            }
            push @full, $line;
            next;
        }

        push @full, $line;
    }
    push @full, '';
}

{ # Process chapters
    foreach my $subfile (@subfiles) {
        my @lines = _readFile($subfile);
        foreach my $line (@lines) {
            if($line =~ /^\=for\ html/) {
                if($line =~ /\<img.*src\=\"(.+?)\"/) {
                    my $img = 'PDL/Book/' . $1;
                    if(-f $img) {
                        my $imgdata = _slurpBinFile($img);
                        $imgdata = encode_base64($imgdata, '');
                        my $imggd = GD::Image->new($img);
                        my ($width, $height) = $imggd->getBounds();
                        my $realline = '=for html <img width="' . $wid
+th . '" height="' . $height . '" src="data:image/png;base64, ' . $img
+data . '"/>';

                        push @full, $realline;
                    } else {
                        warn("Missing image $img in file $subfile");
                        push @full, "Missing image: $img";
                    }
                    next;
                } else {
                    warn("Unknown 'for html' tag: $line in $subfile");
                    next;
                }
            }
            push @full, $line;
        }

        push @full, '';
    }
}

{ # Write complete POD file
    open(my $ofh, '>', 'complete_book.pod') or croak($!);
    foreach my $line (@full) {
        print $ofh $line, "\n";
    }
    close $ofh;
}

if(0){ # Convert file to epub
    open(my $ifh, '<', 'complete_book.pod') or croak($!);
    open(my $ofh, '>', 'complete_book.epub') or croak($!);
    binmode $ofh;
    $parser->output_fh($ofh);
    $parser->parse_file($ifh);
    close $ifh;
    close $ofh;
}
`pod2epub complete_book.pod -o complete_book.epub`;

#{ # Add images to ZIP file
#    my $zip = Archive::Zip->new('complete_book.epub');
#    foreach my $img (@images) {
#        $zip->addFile('PDL/Book/' . $img => $img);
#    }
#    $zip->overwrite();
#}



sub _readFile($fname) {
    my @lines;
    open(my $ifh, '<', $fname) or croak($!);
    while((my $line = <$ifh>)) {
        chomp $line;
        push @lines, $line;
    }
    close $ifh;

    return @lines;
}

sub _slurpBinFile($fname) {
    # Read in file in binary mode, slurping it into a single scalar.
    # We have to make sure we use binmode *and* turn on the line termi
+nation variable completly
    # to work around the multiple idiosynchrasies of Perl on Windows
    open(my $fh, "<", $fname) or croak("$ERRNO");
    local $INPUT_RECORD_SEPARATOR = undef;
    binmode($fh);
    my $data = <$fh>;
    close($fh);

    return $data;
}
[download]

Basically, it concatenates all chapters, and inlines all images. There's a lot more to do to make it pretty and more readable (especially the code sections) and fix some encoding issues. But i do get a mostly usable book with chapters, subchapters, pretty pictures and about 226 pages (default font size on my Boox ereader), including a chapter called "POD ERRORS" at the very end ;-)

Also, i still need to add all the .pl files as their own chapter at the very end of the book, so it's a COMPLETE reference.

Edit: Forgot to say, this is a work in progress, and more posts will follow. But it's a good start for what is, for me, equivalent to a "My First Epub generator" picture book ("with simple, large letters for a first forray into the world of written words")...