With respect to interpreting GPS data. I have a dedicated GPS and that means I download the data into a freestanding desktop app after my trip where I can look at all the data in some detail.
I learned a couple of things. One is that the momentary speed display does not match what the app computes as speed between waypoints. I may be seeing some value on the display and later can't find that in the track data. I think the main effect here is that the saved track tries to economize on storage by coalescing adjacent points that don't differ too much in speed and direction. I even get the sense that this is adaptive: if you try to capture more data, the device will more aggressively throw out existing intermediate data points to make room for the new track data.
That said, I never use any "maximum speed" settings from the device itself, but for bragging rights, I will scan the track for sections that are colored as fast and then look up the speed values. I've never caught anything more than, I think 6.3 knots (or perhaps not even that much). That would represent a speed that is in excess of the hull speed (which is 5.4kn for a DS) but below true planing speeds. Some people call that "forced mode". The
hull speed corresponds to the wave speed of a wake that is so long that the boat always climbs uphill. (As described in the linked post). Now, with the DS hull shape, you can push it a bit faster, into forced mode, and I've seen that happen for long enough stretches that even after averaging out the track point a speed above 5.4 knots shows up.
When you are not sailing in flat water, you can also get that when surfing down waves, because in that case the water also surges forward and your speed over ground will exceed the hull speed for that reason.
The other thing I do is to mount my GPS in a fixed holder, and to not turn it on when not in the holder. That way, I don't get any false readings from moving the GPS inside the boat. If you have visual track data, where you see at what point you record such high speeds, my prediction would be that you would find one very short leg that gives you the peak value, not two or three legs each building to that speed (like you would get when a gust hits).