ClipSpy shows the following clipboard formats:
So the clipboard contains URLs that point to the original image, binary bitmap data of the image and a path to a temporary *.bmp file for the image.
I guess one can not blame MC to take the available bitmap data instead of getting the original files from the URL (as Windows Explorer seems to do).
Therefore I withdraw my complaint. Thanks for pointing me to the ClipSpy app though.
Fun fact: Dragging and dropping an image from Edge to MC works as I would have expected. Why? Because Edge only delivers CF_TEXT and CF_UNICODETEXT in the clipboard and no bitmap data. Like with Firefox, both formats contain the image URL.
About making MC Open Source: I understand your concerns and of course respect your decision.