Thermal: Parts of the detectors need to be very cold. Depending on what you're doing you may have to cool parts of the satellite down to 5 kelvin (-450°F) or even lower. You need to use a cryocooler than won't vibrate too much and impact your measurements.
Power: Each VLA antenna uses about 18kW just for cooling. That's a lot of juice— back of the envelope math says you need a lot of solar panels for that. An area smaller than a football field but bigger than a tennis court. It's manageable if you're only building a few of these, but less practical if you want to build 64 or even thousands.
Positioning: Most proposals for space based interferometry didn't actually consist of individual satellites. They're typically one big satellite with telescopes sticking out on different arms. If they're free-flying, it gets a lot harder to know the precise relative positions of the different receivers.
That’s likely at minimum partially proportional to the size of each VLA antenna, no? Each VLA antenna is larger than any antenna ever put in space I believe.
> If they're free-flying, it gets a lot harder to know the precise relative positions of the different receivers.
That’s exactly what I thought would be one of the biggest challenges, but wasn’t sure if we’d solved that problem or not yet.