This was originally raised in 3239
The jdt.core part has been addressed. The UI, where the rendering to HTML happens has to handle the unicode characters to create the correct HTML content.

The markdown comment is this:
/// Unicode in markdown \u000A///\u000D///\u000D\u000A///here

And this is the HTML created by JavadocContentAccess2.getHTMLContent():
Unicode in markdown \u000A/// \u000D/// \u000D\u000A/// here

Javadoc from JDK creates this:

<div class="block"><p>Unicode in markdown</p>
<p>here</p>
</div>