I hardly see how that is any different, as a simple regex will easily match both. Any further, and we're doing it in a way which we can't convert existing WikiText into tags... And then we're right back at the notion of ditching WikiText without converting it, which as I've already explained is insane.

Then there's the notion of XML structures being cleaner, easier to parse, and not to difficult to read.
And half that is bull.
Sure, a XML language is good for a parser.
But cleaner... That's starting to push it.
And not to difficult to read...
That's just crap. There's no way a pile of <>'s all over the text is anywhere near easy to read, and to someone who doesn't even understand XML that's just bull.
WikiText was designed to be able to be editable by nearly anyone. An XML language will never be clean enough for the average person to edit.
And like I said, don't state any crap about them using WYSIWYG. There are plenty of users who can't use a WYSIWYG editor and will need to be able to edit plain text. And a XML based language isn't anywhere near friendly for them.