I'm a software engineer. And there's a cheeky but widely used principle in our industry, especially among people working with servers (on the server-side): KIWI or "Kill It With Iron". When one has a scalability (hear "performance") problem, one of the options is usually applying KIWI, meaning throwing more server at it.
Where am I going with that? Well... I watched Total Recall last weekend (it is the remake of 1990 movie with the same name starring Arnold Schwarzenegger and Sharon Stone. It is loosely based on the 1966 short story "We Can Remember It for You Wholesale" by Philip K. Dick). Once again "the believability" of yet another sci-fi movie busied my mind.
I think, in Science Fiction, the author (or the director, if it's a movie) has the option to explain something with "advanced science". Drilling a tunnel from the UK to Australia through the earth is amazing. With enough resources and sufficiently advanced technology it is also probably achievable.
However if one has to rely on free fall, it is easy to imagine that 17 minutes won't get you from one end to the other. I'm not going to waste my time on precise calculations, but I'm pretty sure gravity would decrease as the shuttle gets closer to the core and the whole journey would take much longer than just 17 minutes.
Even if we think that the speed is linear and we accept that the passengers can get from one end to the other in 17 minutes (assuming the distance is earth's diameter for simplicity's sake) then this would correspond to a speed of approximately 28,000 miles/hour (45,000 km/hour). Can you imagine blowing a hole through the window, sticking your head out to shoot, climb a ladder etc. with that speed? Come on!
The moral of my complaint is: "Don't Do It If You Can't Kill It With Science". If you want to stretch what's acceptable or believable, try throwing science at it. If it doesn't work, just don't do it.
P.S. I'll deal with a different type on scientific progress in a future post: