Coilguns have physical limitations that make them unsuitable for extremely high velocity applications. Specifically, the engineering difficulty in switching the coils on and off scales with the square of the currents used in a coilgun.
Any inductor has a rate of change of current proportional to the voltage across the ends. So all else equal, if you double the current, you double the switch-on and switch-off times. But, since you're using higher energies, you're getting a faster projectile - which means you now need to switch the coil off faster as well, to prevent suck-back.
So in short, coilguns get significantly harder to build at higher energies. In contrast, railguns get more energy efficient as projectile velocity increases.
I'd guess the main reason is that it's an inherently more complicated design; if you don't sync the electromagnets perfectly you spend energy decelerating the round. it does have the advantage that the projectile doesn't need physical contact with an electified rail, so maybe it could be less damaging to the device itself.
These railguns also have synchronized pulsed power systems that are quite complex. The University of Texas has been researching this stuff for decades. Here's a simple paper I Googled up https://apps.dtic.mil/dtic/tr/fulltext/u2/a639371.pdf
The coils are inductors, and so cannot be turned off instantaneously - you get a voltage across them proportional to how quickly you're changing the current. As you build a faster coilgun, it gets harder to deal with the voltages generated by the sharper current pulses necessary to do useful work. And presumably you're also increasing the current involved to get higher velocity, which also contributes to the engineering difficulties involved in managing the coil voltages.
Amateur railgun builders comment about the projectiles welding themselves to the rails; that certainly doesn’t happen in a coilgun. I think there’s also a big problem with supplying extremely high currents, which you can circumvent a bit in a coilgun by having more turns of wire.
Speaking from ignorance, I would expect that official research would encompass military grade material science, to which amateurs don't have access realistically.
