It's a trap. Drag requires more speed, which creates more drag. So you go with a heavier, more aerodynamically efficient object, which requires a larger railgun. The minimum size for this railgun is probably so large as to make the entire endeavor futile in comparison to rocketry.
Artillery gets around this by firing relatively high. This wouldn't be practical for launch to orbit where you want as much momentum as possible laterally rather than vertically.
Let's say, for the sake of argument, that your space gun's acceleration track climbs the slopes of Chimborazo in Ecuador and releases the launch vehicle at the summit.
From there, if you launch eastward, you get about 460 m/s for free from the rotation of the Earth near the equator. LEO requires about 7800 m/s. Earth escape requires about 11180 m/s from that altitude.
So if you go to LEO, you need another 7340 m/s of velocity. If you accelerate at 3 g, your track needs to be 920 km long. That isn't quite the distance from the end of your track to the Galapagos Islands.
The same length of track could hit escape velocity with constant acceleration at 6.5 g.
For reference, the Large Hadron Collider is 27 km in circumference, and cost about $5 billion. A space gun would be 34 times longer, and portions of it would have to be evacuated and neutrally buoyant at some depth under the ocean. The entire length would need to be powered.
If you built a circular track 920 km in circumference, 147 km radius, the centripetal acceleration required to keep something moving at 7340 m/s on that track would be more than 37 g. That only increases as you reduce the radius. So no getting away with a shorter track.
If you're going to build a megastructure like this, you might as well make it capable of launching manned vehicles. If your typical payload can handle 1000 g, you wouldn't need anywhere near that length of track. You would only need 3 km or so, I think.
A 3 km long coilgun is possible with today's technology.
It still takes a lot of launches before the track starts to be cheaper than just building a new first stage for each launch. It takes a lot more if the first stage rocket is reusable.
I suppose it depends on how fast we accelerate them upto right speed
So if we imagine a Nimrod style dish on top of an aircraft near space, and say orbital speed is 30Mach or about 21,000mph call it 30kph for ease, and minimum useful weight would be 500g we need to input ... 54MW?
And control travel around a dish tighter than anything previously seen.
Yeah - nice idea ...