Possibly two steps: 1) capture screenshot to png, and 2) use OCR app which keeps layout, font size, etc. and can export as markdown. But I suspect that is asking a bit too much. All OCR apps that I have seen export as plain text or searchable PDF (often with an OCR layer).
Edit: I can capture to PNG, open in Nitro PDF Pro (or probably most other expensive PDF apps) which does OCR, and then export as MS Word document. That reduces the problem to doing Word to markdown conversion.