Short answer: choose the file, pick Markdown, copy the result. The notes list any images that were left out and any style the converter did not map.
Structure, not appearance
A .docx stores paragraphs with style names, so the conversion is a mapping from Word styles to Markdown blocks rather than a guess from font sizes. That is why a real heading becomes## and a bold run becomes **bold**, while a manually enlarged paragraph stays plain text.
What does not survive
- Images — counted, not embedded.
- Text boxes, headers and footers — outside the main document flow.
- Comments and tracked changes — not part of the body text.
- Fonts, colours and spacing — Markdown has no vocabulary for them.
Once you have the Markdown, the Markdown to HTML page renders it, and text diff shows what changed against the previous version.