Convenient and tweakable Atom+Markdown+Pandoc+Jupyter experience, export2ipynb

In the proposed tool chain you:

* Write Markdown in Atom (with all available packages that help much),

* Insert Jupyter code cells and deploy via Hydrogen,

* Apply any Pandoc or text filters via [Pandoctools] (https://github.com/kiwi0fruit/pandoctools) to convert to markdown/html/pdf/ipynb,

____* Pandoctools lets you define text filters pipelines ("profiles") and apply them to any file by simply adding two lines to markdown metadata section and opening the file with Pandoctools. You define text conversion algorithm once and then just use it with few 'Open with' clicks (or even with one short key with Atom package).

____* Collecting Jupyter cells outputs is also a Pandoc filter,

____* There are even working cross-references in exported ipynb,

____* Can convert markdown to markdown in order to post-convert with existing tools,

____* Optionally write math in SugarTeX that is a more readable LaTeX language extension and a transpiler to LaTeX.

[Example doc](https://github.com/kiwi0fruit/pandoctools/blob/master/examples): from markdown with Jupyter python code blocks, SugarTeX math and cross-references to ipynb notebook.

See more about all mentioned tools in [Pandoctools repo](https://github.com/kiwi0fruit/pandoctools).

There are also recommendations of nice Atom packages and even good open-source font fallback chains for Unicode-rich markdown documents (monospace for editor, serif/sans for output).