Nothing in life is without risk, and the more we risk the more we live. If we are not capable of, or willing to accept risk we should not participate in anything that requires risk.
Where multirotors are concerned, unless we are using MilSpec components, certified flight controllers, minimum performance standards, post production flight testing of every system, and hardware certification standards we will always be dealing with a great many unknowns. What we fly has absolutely none of those and every time we fly we are a test pilot. We can at best hope things will work as they should, even at the most expensive levels. Just for conversational purposes, the last I heard, which was only a few weeks ago, Lockheed had yet to figure out what to put in the user documentation for their relatively new flight controller and places that have developed a 'self learning" FC don't have a full grasp with how theirs work yet.
I don't know about anyone else but when I obtain a new flying machine
the absolute last thing I do is take them out to fly. I'll spend hours and hours going over everything on the bench, inspecting fasteners, reviewing the control functions, memorizing the locations and function of the flight controls, learning the camera settings, performing calibrations, and all the other little things we need to understand and be comfortable with before we set out to fly. All of that is to mitigate risk. During most of that time the system is powered up, with or without propellers, and has
ample time to collect any GPS data that might be needed for an "almanac". Most decent GPS systems don't need any more time than is necessary to start collecting continuous data from the minimum number of satellites required to generate accurate position reference. For Yuneec that number seems to be 10. More advanced systems require no more than 5 but benefit from more. Sure, the longer a satellite is in view the better the HDOP but as more and more satellites become visible the HDOP also improves until we will won't obtain any more precise data than the entity that controls the satellites want us to have, which in the U.S. is within a few meters and no more without adding RTK or DGPS to the navigation system.
The 920 does not co-locate the compass and GPS with the flight control board, nor does it locate the GPS over a heat and power source. Building a fire underneath the GPS module, which is close to what is taking place with the battery directly under the GPS with the Typhoon H, is a pretty stupid thing to do. That "tower" on top of the body is where that stuff is located which removes it from the most common interference sources. I tend to believe the GPS/compass module used in the H-480 id not as good as it could be (along with most everything else the system incorporates) but how and where they located it is most likely 985 of the problems they have experienced. Where consumer drone manufacturers place their navigation equipment is consistently where the DIY crowd never will. The DIY crowd, those with average or above experience anyway, just don't experience fly away's.
BTW, if GPS is out there's no possible way for RTH to work. If the aircraft doesn't know where it is it certainly won't know where "home"
used to be. For all intents and purposes the system has dementia. Telling it to go home is akin to telling it to "get lost"
Nah... I'd hold to the $12, that was for 2 or 6 sets? 
... hmmm?
I think you're over pondering the matter... In simpler visual terms... On house, if you have a wall socket with 2 hole plug and had a lamp with a 3 prong plug, you need an adapter to interface. If you convert the Home to a 3 prong, you no longer need an adapter. Same principle... use an adapter between the receptacles or change one end to match the other. Really no major pro/cons to it, just a different option. If the battery had EC3 connections you wouldn't need adapters on either H920 or H920 charger. 
