A 920 can be seen well enough to operate without having to use the camera view a little over 2000’. Bear in mind a 920 is ~1.76x larger than an H Plus or 520. Having built and flown 1200mm multirotors I know they can be seen and flown without need of the camera view out to 2500-2700’. A 14’ (4267mm) diameter multirotor becomes unmanageable without needing the camera or autopilot a little over a mile. Note I mention not using the camera or autopilot above. It’s all manual control.
For as long as I can remember one of the most common questions asked by visitors to RC flying fields has been “ How far away can you fly it?” Other common questions are how high, how fast, and how much but those aren’t relevant for now. The answer to the “how far” question has always been “As far as I can see it.” Between then and now nothing has changed. With model airplanes that distance has been somewhat farther than can be done with a multirotor due to aircraft design. An airplane has distinctions that permit orientation much further away. Long wings in front, small wings in back, a vertical rudder, all assist maintaining direction and orientation. Fixed wing provides a defined front, back, and vertical orientation that a multirotor does not. But even they have visual distance limits controlled by size and perspective. The larger they are the better they can be seen as distance increases. A 10’ wing span airplane cannot be seen well enough visually at 1 mile to effectively control it even with 20-15 visual acuity. I use those numbers because I have that vision along with four 10’+ model airplanes and can use actual experience to make reference.
Model aviation shares much with full scale aviation. Flight and route planning, obstruction clearance, safe operating altitude among them. Flight planning includes allowances for “what if” situations. What if the engines fail? What if the electrical system fails? What if the fuel supply becomes too low?
As multirotor operators we need to add; What if the autopilot fails? What if the camera fails? What if RTH fails to function? What if radio or video signal attenuated or experiences interference? Will I be able to see it well enough to fly it manually? Is it high enough to maintain clear radio signal? We should never be 100% dependent on automated systems to effectively fly the aircraft as automated systems can and do fail. It’s for that reason they still have pilots on commercial airliners. People have already died because they depended on automated systems in expensive cars, not to mention private and commercial aircraft.
There’s a lot to learn with aviation related activities to do them safely. Safety, both for ourselves and others, is the foundation of model aviation. Any time we cannot assure the safety of people uninvolved or unaware of our modeling activities we are acting irresponsibly and putting them at unnecessary risk. If what we do as individuals could only effect us as individuals we could do whatever we wanted to but that’s not how it usually works. What we do has the potential to effect a great many people, including the entire population of our hobby, so we must shift our thoughts from what I can or want to do to what can I safely do? How will I be able to mitigate a situation if and when something happens? We must employ mature thought processes, not impulsive ones.
Having been involved in model aviation, commercial/military UAV’s, manned aviation, and multirotors for 48 years I still don’t know everything yet and learn something new every day, but during the entire time I’ve been involved with that stuff safety, both for myself and for others, has been a mandatory part of all of them.