>>Heinlein was among the worst in seeing technological trends

Agreed, but he had his moments of prescience - notably Waldo, Inc (those remote manipulators called 'waldos' are named in honor of the story - which was partly about the invention of remote manipulators - only Heinlein's had force-feedback, something we don't have yet) It was also about potential issues around widespread wireless - broadcast energy (your car needs an antenna and a motor, not gasoline and an engine) - it was making people sick but the corps selling the energy acted like tobacco companies were at the time of writing (whether Heinlein was aware of that or not).

More generally, I'd agree that science fiction can be pretty terrible, but disagree in that it can be terrible in that the characters act in totally implausible ways (EE Smith, Verne, Asimov). And, yes, Ian Banks does a great job - as does Nathan Lowell (another fan here). Also, pretty much anything by John Scalzi contains very believable behavior.

Oh - back in the 70s, Hal Clement, author of the Barlenan stories, and who was "Harry" to me, asked for a way to use his PC to calculate orbital mechanics, because he was killing himself getting the science right. Turns out the PCs of the day just weren't capable with anything off-the-shelf - but it started me on scientific programming languages.