> I can't follow your reasoning at at. "Drop the height h" is completely ambiguous.

I'm referring to the classic proof where you drop the height perpendicularly to the hypotenuse from the opposite corner.