... it is not inflight-removable, but leaves no traces when detached (on the ground).
Not inflight-removable=permanent. Whether it can be detached on the ground without leaving traces or not does not make any difference. Our spraying devices were also ground detachable without leaving traces but still required a full certification program.
... when they have zero influence on the aircraft performance ...
How can you know without an analysis and test program?
... the test data (by RAM Mount) is well above the aircraft performance?
Whose test data? Obtained how?
Sorry, I cannot write anything different, because beside being a pilot I'm also an aerospace engineer...
Outside is a bad idea. Depending where it is mounted it can have a lot of influence on flight characteristics. I've seen it on top of the elevator surface of an old PA28, under the wing.. I would not do anything like that. Can you guarantee that stall characteristics are not affected? If yes you can show us your STC for the camera.
Can you guarantee that stall characteristics are not affected?
Much more problematic is flutter, which can be provoked by almost anything. On YouTube there is a video of a Hawker 800 with an approved (!) installation of winglets that under certain conditions led to oscillations. The guys are lucky to be alive.
I've seen it on top of the elevator surface of an old PA28...
Incredible. Mounting anything on the all-flying elevator of the aeroplane comes very close to committing suicide.
I think outside is a bad idea, unless really well secured (which most aren't) though "everybody" is doing it.
If it comes off it could do a lot of damage - to your plane or the one behind. One decent dent in an elevator could be a four-figure repair.
My understanding is that getting it approved (for an ICAO CofA aircraft) is a long process so nobody bothers to do it legally. For a homebuilt, I don't know the rules. LAA in the UK?
I am not aware of an insurer not paying out but presumably those who crashed but survived (and they usually did have to survive for the footage to end up on Youtube) removed the camera before anybody turned up.
One can get decent results with a camera inside the cockpit, provided it is behind a clean scratch-free window, and it's much safer.
This proof of concept
(rather boring) was shot looking forward. The antivibration mount used is here.
I spent some time trying to find a way to mount a "bullet camera" (which is much more aerodynamic than the brick-like GoPro) outside the cockpit in the TB20. There is a safe way to do it, involving the huge suction cups on the inside of the door, and a ~1.6mm aluminium strip, about 100mm wide, going up and outside from them, through the rubber door seal, to a camera located immediately outside the top edge of the door. I gave up on this when a Sony "semi pro" camera I was going to buy (for about €800) turned out to have been discontinued.
As another opinion 
I think you can tell a movie made with a GoPro, in the same way that in the 1980s you could tell when somebody used Pagemaker to do a brochure for some product... The raw image quality is good but the image distortion gets very tedious after a short while. And prop effects are very hard to avoid. The same money spent on a camcorder will deliver a much better result, IMHO.
in the same way that in the 1980s you could tell when somebody used Pagemaker to do a brochure for some product.
and Corel Draw in the 90s. lol
Not inflight-removable=permanent.
Thanks Whatnext for clarifying that. That's the information I was seeking. No test data or test program available or done. I scanned the websit of RAM Mounts and GoPro but couldn't find any information on this subject.
And thanks mdoerr, I personaly also think it has some influence on the performance but whether you will notice it during flying depends on a lot of factors, like the one you mentioned. A lot of GA pilots who are using the GoPro's outside the aircraft say the don't feel any difference in the performance. But than again, no test data. So, no STC. ;-)
I think a small item, fixed to the hull of a slow aircraft, is not going to make a noticeable difference. It will create a bit of drag - maybe 1-3kt on a 100kt aircraft.
I would absolutely not attach it to a wing, elevator, or any control surface.
The problem is how to fix it. Forget suction cups. People have attached them to Cessna wing struts with brackets, which should be OK if done correctly.
But to do it legally would be a long job because you have to go through a flight test programme. Pilot-DAR here might have a few things to say, as he does that for a living.
As I said, try to do it internally.
Also I would add that if you can get a camcorder with a manual shutter speed setting, setting 1/100 or slower removes prop effects nicely.
Nothing wrong with Corel Draw - I still use v11
Nflightcam sells a "spud" which can be permanently screwed to the airframe, on top of which you can (more or less) securely attach your GoPro.
And prop effects are very hard to avoid.
There is a lens kit available for the GoPro which filters out the prop distortion: NFlightcam Aviator Lens Kit for GoPro
Someone I know (ahem... err) attached a GoPro to the wingtip skid of a Bergfalke 4 glider (which has an 18 metre wingspan, much longer than most GA aircraft). Absolutely no effects were felt in the glider. If a slight amount of rudder pressure were needed to keep the yaw string in the middle it wasn't so much that it could be percieved. Similarly, a Go Pro strapped with the standard strap mount to the struts of a 100 mph GA aircraft could not be detected by the pilot. Of course, if I ever put a GoPro on the Auster it's in flight removable (even if external). Perish the thought that I ever do anything else...
On the other hand, the same camera mounted on the head of the pilot turned out to be a much bigger flight issue due to the distraction it caused.
The thing I hate about GoPro and its ilk are the resulting artefacts seen when looking through a propellor. It doesn't seem that there are small camcorders with 1080p resolution that can be mounted externally that do not suffer from this. I'd love someone to produce a GoPro sized 1080p camera with a CCD. The old (SD analogue) Sony bullet cameras were much better in this respect.
