Great question - the prop systems are designed to inevitably operate in vacuum, so we ground test at atmosphere first to verify the hardware performance and then test the system in a vacuum chamber to simulate space.
[edit]: Both options of thruster choice make sense for different missions so it starts with customer conversations usually. We'll work with a customer to understand their mission requirements and then design a system to match. Some of the things feeding into the trade would be the mass of the spacecraft, how fast the customer needs to get to orbit, the footprint we can take up externally, and any preferences for redundancy. I'll also look to see if I can build a system with existing thrusters vs. designing a new one. (For example, if the customer wants 48N of thrust, I would suggest using 2 x 22N thrusters and getting close to the goal instead of designing a custom engine)