There's a lot of posts about MarkDown and TW5 at the moment, so I'll reply here with as much detail as I can, and then review the other threads.
The situation is that:
1) TW5 wikitext is currently a blend of:
a) classic TW syntax for most basic formatting
b) MarkDown syntax for a few bits of basic formatting (code blocks, and allowing HTML tags)
c) GitHub Flavoured MarkDown for a few other bits of basic formatting (strikethrough, fenced code blocks)
d) new syntax for TW5's tiddler operations: transclusion, lists, widgets, macros
2) My declared aim is to simplify things by adopting MarkDown-esque syntax for all basic formatting syntax, extending it with TW5 specific syntax for tiddler operations
My motivation in adopting (more) MarkDown is to take advantage of its familiarity to many users, and to permit the kind of interoperability that Yitzhak describes.
I'm familiar with MultiMarkDown. Much of MMD is oriented around page media (footnotes etc), and isn't directly relevant within TW5, but there's definitely some interesting stuff there. I'm not sure what I think of MMDs table implementation.
So, I think that the best way to interoperate with Scrivener will be by restricting content tiddlers to a common core of MarkDown syntax. If one wanted to publish that content both via TW5 and Scrivener, I'd expect to have to wrap it in separate templates for each target.
Best wishes
Jeremy