I'm very glad that this company exists, but this article doesn't help its case very much other than "Hey look, they got a load of funding".
Elon Musk wants to put 12,000 satellites into orbit. It would be very strange to me if he doesn't have someone looking into the feasibility of a giant subterranean launch tube to do something similar. Get the payload up to Mach 10 on terrestrial power and complete the orbit with a small single stage.
Hyperloop?
A launch tube would be very different, if for no other reason then the fact that you need to actually jettison the payload at one end at high speed.
"Tom Swift and his Subterranean Launch Tube"
- get it high out of the atmosphere - give it a sideways kick so that it's going fast enough as it falls back that it misses the earth
Going straight up minimises the velocity lost by drag, but means you have to pull a right angled turn, going straight sideways grossly increases drag but makes the energy for insertion less
A railgun can only put one velocity vector on a launch, you still need a rocket of some sort to pull that turn
On the point of 3,000 MPH. One of the comments on the TechCrunch article maybe provides some insight on why they aren't targeting 17,500 MPH.
"even if they need to supplement with rocket propulsion (after exiting the rotational acceleration phase), the amount of rocket energy needed will be way less. could see a huge increase in payload mass fraction as a result. maybe closer to 20 or 30% instead of 1 or 2%"
IF they are flying beyond Mach 1, drag coefficient will actually drop as Mach number increases. [1] Gravity loss will play an interesting roll as well for a system launching with a high initial velocity. There will be a balance between aerodynamic drag and gravity drag losses. The lower the angle launched, the higher the aerodynamic drag but the low the gravity drag loss. [2] Typically launch vehicles actually need more than 9.5 km/s of delta-v even though the orbital velocity for low earth orbit is only 7.8 km/s. This is due to predominantly to account for the large drag forces the velocity experiences during the vertical portion of ascent at sub, trans, and lower super sonic speeds prior to exiting the atmosphere. Some allocation will need to be made for drag loss, but I suspect it is less than a traditional launch vehicle. I would also suspect there is a balance between launch speed, fuel fraction reduction, and difficulty of implementation. The sweet spot between those three things will be important to hone in on.
[1] https://en.wikipedia.org/wiki/Drag_(physics)#/media/File:Qua... [2] https://en.wikipedia.org/wiki/Gravity_drag