I'm pretty sure it's just the way that Markdown runs through things.
[Should HTML comments be stripped when validating answer length?](https://meta.stackexchange.com/questions/25344/should-html-comments-be-stripped-when-validating-answer-length)
will become
<a href="https://meta.stackexchange.com/questions/25344/should-html-comments-be-stripped-when-validating-answer-length">Should HTML comments be stripped when validating answer length?</a>
first, which creates this:
<!-- **Possible Duplicate:** >
<!--<a href="https://meta.stackexchange.com/questions/25344/should-html-comments-be-stripped-when-validating-answer-length">Should HTML comments be stripped when validating answer length?</a> -->
Then the Markdown parser starts removing HTML that it doesn't allow:
It removes <!-- **Possible Duplicate:** >
first.
Then it sees <!--<a href="https://meta.stackexchange.com/questions/25344/should-html-comments-be-stripped-when-validating-answer-length">
as one unsupported element and removes it.
Then it removes the </a>
.
And you're left with Should HTML comments be stripped when validating answer length? -->
.
Edit: As explained before, Markdown just goes from <
to >
. It doesn't look at comment syntax.