Currently Liferay generates an hreflang alternate url in the head of every page for every translations available for that site. While this is helpful if you entire site is fully translated in every language, it becomes a problem when not every page is translated.
With content pages, we need a way for Liferay to know if there is a translation for a given page and only populate the hreflang if there is. Also, that url wouldn't even exist if there wasn't a translation.
Without this, site crawlers scrape all of the untranslated pages and flag some of them for duplicate content.