Where is the sensor on the nv5, if it's before the escapement stage, then it can't have trouble with trills, if it's after the escapement, and the action is not properly regulated, then trill problems can happen.
The sensor rail detects a small "wing" or "fin" that sits on every hammer butt. It passes through the detection point(s) as the hammer butt rotates (with the hammer flying). I posted this photo last year, where you can see it:
I have not examined the sensor rail, but I suspect that the same optical hardware works in two phases. First, it detects the hammer velocity (on key-down). In the second phase it detects key release (without velocity information). To do the latter, the backcheck needs to hold the hammer in a position that still covers (one of) the sensor(s) in the rail.
Since there is no measured note-off velocity information, I suppose that it's the last sensor of the return travel which remains covered while the key is kept down (IF there are multiple sensors).
On my NV5 I found that the note plays right before the hammer actually escapes. I don't know if this is an inherent requirement of the sensor rail or if it's only how my NV5 happened to come from the factory. I need to release the key quite a bit before it detects the release (note-off). Without having done exact measurements, the let-off completes maybe at about 75% down, and note-off happens at about 50% up or so.
These are the mechanical details. I'm curious what you'll extrapolate from them.
About the decision NV5 vs CA99: If I didn't have either, I'd clearly go for the NV5. See the thread that I initiated about it to know more about why. I considered the CA98 at the time. Note that I have the NV5 for over a year now and it keeps motivating me to practice every day.
BUT since you already have the CA99, it really depends on how much you "need" accurate acoustic touch. Some features (very few, but still..) are better on the CA99, so you're not 100% upgrading with the swap. Unless you're all-around unhappy with your CA99, you should really try the NV5 before committing. For example, the plan of "fixing" a potentially heavier action with software settings is contrary to the idea of getting a real (!) acoustic action. In the end you might just have different complaints after the swap.
To me personally, the most important differences would be:
- The weight of the keys when they are already down, i.e. the force that my fingers have to apply to keep them down. Should be low so that I can partially relax the fingers.
- By extension, the force profile that "fights" against my finger while it travels from top to bottom. It should be mostly equal during the travel and low near the bottom (after escapement).
- Is there remaining inertia that I need to overcome when I want to repeat a note shortly after playing it already? On digitals, you can observe "key bounce" when you release a key. The more and the longer they bounce, the higher the impact of it for playing repeated notes.
I haven't tried any CA9x recently, so I can't tell. But it's night and day on my old CN37 vs the NV5.