Over the years I've seen a push and pull between actual published science fiction writers and artists and those who think more science should be involved in sci-fi story telling. There's been a hugely failed push by some to get sci-fi to really go hard science across the board.
Two of the best examples is the bad movie science websites and blogs and the never-ending "No Stealth in Space" argument.
Neither has gained any real traction. You still get gobstoppingly nutty concepts like "Red Matter" in Star Trek 2009, and Cylon angel time loop, parallel evolution in Battlestar Galactica.....and Ancient Aliens is in season, what, three? Four?
The reason is simple: People don't expect it to be spot on. All film and writing is about story. Its about people. The machines are ALWAYS backdrop. There only needs to be enough science to give the viewer or the reader permission to enjoy what they are watching. The requirement of suspension of disbelief changes from viewer to viewer, but you are safe going just a bit above the science knowledge of the average population. Aim for the disinterested college graduate and you're good.
Even the Hard Science sci-fi, like 2001 often sets you up for the patently absurd, like monoliths and LSD trips through space-time. By the time you get to the drug-laced monolith trip, you can forgive just about anything after two hours of apes, people walking in slow mo in zero G and rotating rings.
Margaret Atwood, writer of The Handmaid's Tale, once wrote that "if it is realistic or plausible, it isn't science fiction."
I wouldn't go that far. I'd say you need enough to give the viewer permission to enjoy the story. And the truth is the better the story, the better the immersion, the less need for science. By the time most people got to the floating mountains in Avatar, they didn't give a **** if that made no logical sense whatsoever. They were just like "floating rocks! Bring it! I want more alien forest flying beast 3-D!" Had they tried that in a movie with a weaker story, the response would have been far, far different.
That decision on where the line is will not be governed by science, or scientists, or fans of hard science. It will be governed by the fact that J.J. Abrahms Star Trek grossed more than $385 million in foreign and domestic sales, despite Red Matter, mighty morphing Enterprise interior scaling and transwarp beaming or whatever the heck that was.
So how much science do you need to enjoy the story? How much handwavium is too much? How much of it matters if the story and setting are engaging?