Re: Makes you wonder
It's pretty easy to work out that won't save you much.
Suppose you accelerate along this tube at 2G. Then in 10 seconds you have covered nearly 1km; that's a big tube to build. In 20 seconds it's nearly 4km. You need enough linear induction motors or whatever(*) with sufficient power to accelerate a hundred tons at 2G for that whole distance - also lifting against gravity (depending on what angle you point this tube at)
However, after 20 seconds you've reached less than 0.4km/s, only about 5% of low earth orbit velocity.
Even if this saves you say 10% or 15% of fuel load (I haven't worked it out), the rocket itself is almost the same size and just as expensive to build. So the payback per launch is very small, for a massive piece of infrastructure.
(*) You could use auxiliary rockets instead of LIMs. But in that case, you might as well attach them directly to the main rocket, and you've just invented booster rockets. No tube required.
You can do the math for non-human rated loads too. Let's say you can accelerate small payloads at 10G. To get them to 4km/s, which is still only half low orbit velocity, takes 41 seconds by which time you will have covered 82km. They'll still need to fire their own engines to do the remainder. To place them directly into orbit at 8km/s, your super powerful railgun needs to be 392km long.
Biting the hand that feeds IT
