The problem with super-resolution is that it assumes you're working with a 2D image and what you get is a two dimensional image. It doesn't understand depth and you lose it as a result of searching for more "detail". The image of Ben Kenobi is a good example:
http://screenshotcomparison.com/comparison/121686
His face and hair just look flatter and so does his robe. You lose meta-detail in search of micro-detail.
And in motion it just looks worse as is "carries over" the "detail" frame for frame, transporting flat "detailed" images over the depth present in the source.