Why would it NOT more risky to fly at night? Even for a tested airplane, flying at night is more risky due to limited visibility and need to rely on instrumentation. In a unexpected situation such as a sudden burst of wind, that will be critical.
The best you can come up with is wind?
And what you are describing is only applicable to inexperienced novice pilots. Test pilots are anything but novices and would be able to fly instruments only at night as easily as they can in daytime.
Testing aircraft is even worse, because if everything is expected, then you won't need testing. Take off and landing due to unexpected events or damage will be much harder due to limited visibility. In air, it will be much harder for the chase aircraft to inspect damage, unexpected vibration/flutter due to limited visibility.
Know many kinds of birds that fly at night instead if sleeping? That is the most likely cause of damage that can be sustained in the air, the only other remotely likely scenario would be a collision with a chase plane.
Unexpected vibrations/flutter does not develop suddenly and without reason. If the plane flew fine during daytime, it will fly the same at night provided you don't do anything you haven't tried out before.
All of your examples are far fetched to say the least, and already debunked by the fact that the F117 did most of its test program in the dark. It is a fact, not theory. What more, no-one is saying the J20 is going to such lengths. However, if the plane really has already had at the very least developed a basic flight envelop elsewhere before being showcased at CAC as I have been suggesting all along, it would not be much of a risk to fly the plane in at night for the 'first flight' display, and then fly it out again under the cover of darkness to a proper fly test facility where the real work could continue.