[Mr_Minderbinder]

100%, this is the fundamental problem with all these techniques. We don't treat our other, more established, historic media/artefacts with the same contempt. If an ancient text fails to mention the colour of the emperor's robes we do not "interpolate" the text with our own ideas of what it should be in our new editions. This would be universally recognised as a textual corruption yet when photo "colourisers" do the same to colourless photographs they think they do the world a service. Why is historic media interesting and valuable in the first place? Fundamentally it is because a historic artefact conveys information from that time period. This is why none of these hyper-restored versions interest me, they are far too corrupted by false or spurious information. Time corrupts them enough already, we don't need to add more. This is the modern version of historical embellishment/exaggeration, contorting it into a more palatable or appealing semi-truth instead of telling the uglier actual truth. The motivations need not be nefarious, they could be entirely commercial like it is today.

They are more tools of embellishment than tools of restoration. A good restoration actually "restores" information. If an old manuscript is missing a page or contains an error, we replace it with the same portion from another manuscript. Likewise if there is a scratch on this frame we should use the same frame from another print to patch it over. We should try to develop our tools to do more of that sort of work. A motion picture produced in the 1990s does not need to look like one produced in the 2020s, in the same way a Rembrandt does not need to look like it was painted in Photoshop.