Support reading of title from H1 element

The Markdown module can substitute `title` value by the value of the first H1 element present in the body of text. That makes the Markdown document more generic and less bound to the particular reader. Would it be possible to extend `myst-reader` to be able to do it as well, please?

Start by locating the myst-reader path that currently determines a document's title and inspect how the first H1 is exposed by the Markdown/MyST parser. The work is done when a document without an explicit title can use its first H1 as the title without changing documents that already provide one.

Written by the indexing model from the issue text.