All I can tell you, I experienced directly the recorded value of 50+ on the Ambient, and the speeds coming off that last storm were much greater. Not empirical and seat of the pants, but I'm telling you...
I do understand, but eg were the Ambient and VP2 at exactly the same position and height? How sure can you be that the Ambient was correct? It's a common experience that despite strong winds, an anemometer may be in a partly sheltered environment if it's at all close to buildings, trees etc and may appear to be under-recording. It's a shame we don't have 3D ways of looking up into the air and seeing how the air currents swirl and blend due to ground friction, obstructions like buildings etc, but all we can realistically do is to get the anemometer up into clear, free-flowing air as far as possible.
The thing about cup anemometers like the VP2 one is that there's not too much to go wrong with them that would make them under-record other than poor exposure. Provided the cups are spinning freely on their spindle then the speed of rotation is determined purely by the geometry of the cups, which obviously isn't going to change over time. And, certainly for the VP2, this has all been checked in wind tunnel tests.
Other than that, the only other possible source of error is under-counting of the revolutions. In older 6410 units with a reed switch then a faulty switch certainly could miss revs, but for a solid state switch in anemometers made in the past 4-5 years that's unlikely now. I guess the counter circuit on the SIM is another possible source of error, but again that seems pretty unlikely - these things tend either to work correctly or fail completely.
So, my money would be on limited exposure.