Shall CommonMark inline inside HTML5 render as CommonMark? If not, why not?

At github.com/MALSync/MALSync/issues/2092:

…via the available “GitHub-Flavoured Markdown” markup syntax, I am able to utilise CommonMark inside a <code> tag, directly inline:

1.	<code>[por](https://www.loc.gov/standards/iso639-2/php/langcodes_name.php?iso_639_1=pt)[-](https://stackoverflow.com/questions/13630355/culture-name-using-underscore-instead-of-dash#comment18695296_13630355)[BRA](https://en.wikipedia.org/wiki/ISO_3166-1_alpha-3#:~:text=Plurinational%20State%20of%29-,bra,-Brazil)</code>;

	<!-- 
	
	[^1]: [`stackoverflow.com/revisions/40076454/2`][1]

	[1]: https://stackoverflow.com/revisions/40076454/2#:~:text=Sometimes%20you%20need%20to%20encode,%5BSyntax%5D(http://en.wikipedia.org/wiki/Syntax_(programming_languages%2529)

	 -->

1.	<code>[por](https://www.loc.gov/standards/iso639-2/php/langcodes_name.php?code_ID=363)[-](https://stackoverflow.com/questions/13630355/culture-name-using-underscore-instead-of-dash#comment18695296_13630355)[EEE](https://en.wikipedia.org/wiki/ISO_3166-1_alpha-3#:~:text=ABB%20Asia-,eee,-Europe)</code>;

1.	[`ro`](https://www.loc.gov/standards/iso639-2/php/langcodes_name.php?iso_639_1=ro);

1.	[`sv`](https://www.loc.gov/standards/iso639-2/php/langcodes_name.php?iso_639_1=sv);

1.	<code>[zho](https://www.loc.gov/standards/iso639-2/php/langcodes_name.php?code_ID=)[-](https://stackoverflow.com/questions/13630355/culture-name-using-underscore-instead-of-dash#comment18695296_13630355)[HANS](https://stackoverflow.com/a/4894634/9731176)</code>; and

1.	<code>[zho](https://www.loc.gov/standards/iso639-2/php/langcodes_name.php?code_ID=)[-](https://stackoverflow.com/questions/13630355/culture-name-using-underscore-instead-of-dash#comment18695296_13630355)[HANT](https://stackoverflow.com/a/4894634/9731176)</code>.

I was under the impression that CommonMark requires a newline between the start and end tags of any HTML5 for the CommonMark within to render, which I still believe to be correct. Please correct me if I’m not. However, with that premise, if this is technically feasible, why is it disallowed by the standard?