Determining if something is poorly formated or not is a very difficult problem. Determining if they have pasted code but not code block formatted it is a much bigger problem than stray
<tags> in the text. Some of which are whitelisted (please look at the source).
Trying to write code code to determine what the intent of the text - did you want to use
<i> there? or was that supposed to be inlined is a non-trivial problem.
Human meaning is best determined by humans who read the post and assist by demonstrating the proper formatting along with instruction on it... rather than having some code attempt to guess the intent of the author.
There is no coded solution to people posting questions or answers who don't care about the quality of the post.
<some tag answers="http://meta.stackoverflow.com/revisions/212187/1">
No. You should be allowed to do that because the alternative is painful.
The alternative involves trying to parse the body of the post to determine if something is rendered or not. This isn't exactly trivial. Identifying what is rendered and not means running the post through markdown and examining the rendered output for displayed character counts. Such things as code blocks and Mathjax makes this very non-trivial.
Just consider this answer and the processing necessary to determine if this is rendered or not.
Then you've got things that look big, but are similarly small. Just a guitar note on the music site, or a simple mock up on the ux site, or a single chess position on chess.SE. These are all things that take more than the minimum character count, difficult to render and determine if it is something to say "you need more material for"
Honestly, I'd be more concerned about filtering answers where the entire thing is a code block than things that use html/markdown trickery to determine if its rendered or not.
Votes will sort out if its good or not. Edits will sort out when there's a mistake in the user's understanding of markdown (you did leave a good edit message, right?).