I don't have a lot to add except I inadvertently flew a fixed wing drone (xuav talon) in icing conditions one time. I can share a few comments/observations from that experience.
1. I didn't realize it was icing conditions aloft. On the ground there was a really fine, very light, intermittant drizzle that was not freezing. Visibility was fine, winds were dead calm, it's hard to schedule multiple people for another day, we were all out there, so we decided to go for it. The temps on the ground were just above freezing though (I think this was an early January, Minnesota day.)
2. The flight lasted about an hour, I didn't notice anything amiss from the ground station data or the aircraft performance during the flight.
3. The area we were flying didn't have a great place for an autonomous landing (fixed wing approach needs a lot more room) so the plan was to take over manual control at the end of the survey and land manually.
4. When I took over manual control, the plane immediately dove. I could tell something was pretty far out of whack, but at this point I didn't know what, and didn't have time to think too much about it. I was able to mostly compensate by holding a *lot* of up elevator. The approach ended up being super fast and low, but I was able to spot the landing ok at the very end. I'm not a super pilot, but I guess ok enough when it counts.
5. I walked over to pick up the airplane and immediately realized what was going on. The front 1/3 of the wing was sheeted in ice. The nose was covered as well. It wasn't really visible, but you could feel it of course. This added weight (all ahead of the CG) and so completely threw off the CG and made the plane extremely nose heavy. I was super thankful my pitot tube didn't freeze over because that is a critical part of the speed/altitude control loops (total energy control system.) The ice quickly melted off in a few minutes once I was back on the ground.
Conclusion: I feel super lucky that the plane survived the experience, and I was able to rough out a smooth manual landing at the last moment. I didn't even consider the conditions could or would lead to ice build up. I feel like I learned something that day. In the future I wouldn't send the airplane if the conditions were similar. Knowing what I know now, it would be a poor decision. Knowing what I knew then ... I'm just glad I was lucky to not have a mishap.
I realize everyone else here flies multirotor, so probably they behave a lot differently, and would spend much shorter hops in the air ... so maybe the decision cutoffs would be different compared to a fixed wing? Still, I think it would be a pretty poor decision to send any drone into known icing conditions.
If anyone is interested, back at the time I made a boring youtube video showing the stitch of the imagery from that flight using our in-house designed map stitching/exploration tools: