10k G should be doable, during WW2 the allies had radio proximity fuzes for artillery shells that exceeded that: https://en.wikipedia.org/wiki/Proximity_fuze
Pressure and ballast balls.
Most satellites are unique, so the same can't be done. Besides being weight critical, every failed satellite becomes space junk.
With regard to weight, they say this system works best with very dense payloads. But that's only for the spin/toss stage; it doesn't account for the rocket booster needed to actually bring the satellites into orbit.
As for testing, engineers don't need as much empirical testing anymore, since we have very good computer simulations now. Most of the satellites launched with this would probably be variations on a handful of designs. But I don't understand where the demand for such a launch system comes from anyway. There is only so much demand for school project cube sats; communication satellites launched this way seem dubious (not least because you can launch dozens/hundreds of them at once with a rocket like Falcon 9 or Starship)