The sensor wouldn't wait for the whole travel and hit of the hammer, instead it would measure hammer position/velocity at all times and replicate that, so the latency could be fairly small, I would be curious if small enough for a musician.

In the article they mentioned "[...] activated by solenoids, a proven technology even used on some Spirio Steinways.". On Spirio Steinways's website they advertise a player piano model called Spirio [0], and in the description it they mention their capabilities as "[...] measure hammer velocity (up to 1020 dynamic levels at a rate of up to 800 samples per second) [...]", a sensor at 800Hz would mean a sensing delay of 1.25ms as a starting point. To that we would need to add the mechanical delay of the solenoid, unfortunately I couldn't find a source for how fast are solenoids (the fastest I found was a solenoid hydraulic valve fully closes in 13ms, but that is the full travel, I couldn't find info on response time for first movement), so I'll make an out-of-my-ass guess that a fast solenoid starts acting in two or three ms, then round up and estimate the total delay to 4ms.

So would my 4ms guess be too much for a musician? I don't know, not my fields, my guess (I've been doing that a lot today) is that if we are talking tempo it is probably too little to matter. So my only worry would be phase, if a musician's ear is sensible/trained enough to detect phase (I don't know if it is, probably isn't right?), then a 4ms delay would mean being in antiphase to the expected response at ~125Hz, that is the B2 key.

[0] https://br.steinway.com/pianos/steinway/todays-steinway/spir...