what a hideous overcomplicated css
by far the quickest fix is to regex out all the <a... page Id...> stuff. book will work fine without it
the code i use in Sigil - which removes all <a > entries - is find
</?a ?([^>]+)?>
replace all with nothing
my alternative lazy option would be an epub to epub conversion in calibre which hopefully will tidy up the mess and either fix it, or make it more fixable