Thoughts from a non-lawyer.
Digitally enhancing light levels, (and I assume to some extent by default color temps, hue, saturation, freezing images, etc), has been ruled admissible in this case, http://64.233.167.104/search?q=cache...nk&cd=10&gl=us and a US circuit ruling is strong precident.
But I think enhancing resolution depends upon the pupose the evidence is intended for, as it doesn't just modify existing pixel data - it ultimately creates it. Granted the methods used could present some high probablilities of the enhanced image being pretty accurate. And courts will likely rule that unless there is "...clear and prejudicial abuse of discretion", that evidence may be admitted. But in the end, evidence admitted must then be weighed as to its credibility. If you double the resolution, a lawyer is going to turn to the jury and state that 75% of the pixels in the video they just saw of the person the prosecutor claims is the defendant simply never existed prior to enhancement. They were created by a computer program and/or human hand. Now you can explain until your blue in the face how its done, how the algorithms work, the probability of error, etc. But in the end it's still a "doctored" video, not just light enhancement, etc. And there is going to be a lot of very typical resistance to technology, suspicion, and in many people's mind perhaps some reasonable doubt.
And its biggest potential use presents the biggest problem - trying to use it as evidence as to the identity of a defendant. Without further ID by an eyewitness or other evidence placing a person at the scene, I have doubts as to it being able to hold up on its own until established by lots of independent testing, and then of course by case law. Good perhaps for corroboration, but not a substitute for a positive ID by an eyewitness.

