Is there a difference between necessary and unnecessary bad science?

Classic science fiction is full of stuff that makes no sense, from a scientific standpoint. And it's also full of cliches, which were bright and shiny once upon a time but now just feel pointless. But is some of that stuff actually necessary to the genre?

The Center for Science and the Imagination over at Arizona State University has an interesting blog post about doing away with tired old conventions in science fiction. And along the way, Jason Krell suggests that maybe some unscientific or overused ideas are necessary, while others are just fun:

A good example of a necessary convention is faster-than-light travel/communication. In a story that has a lot of space travel it’s the only way a character can get anything done in their lifetime without sacrificing all connections to their family, friends and colleagues. It will almost always be necessary if the author wants to get out of the solar system.

In other cases, it’s a bit less clear cut. An author has to weigh the cost of staying true to science with telling a compelling and dramatic story. Despite the fact that having a dogfight in space would be the stupidest form of space combat, Star Wars wouldn’t have been quite as emotionally engaging without the tense fight around the Death Star.

But how do you decide which notions are essential to a good space yarn (like a way to travel faster than light) and which are just kind of nice to have, but maybe should be phased out? That's the tough part. [ASU]