Different pianos, different tuning.
It is well known that our physical modeled sounds try to "clone" the original instruments as close as possible. Part of this is also the tuning pattern.
For example, talking about the Seven and the physical modeling of pianos and electric pianos, depending on the sound selected, the tuning (increasing or decreasing on high or low) will be different on every sound.
The reason is because we are just "cloning" the original instruments. Additionally, there's no way that 2 different instruments will act differently, we are talking about digital instruments, there's nothing there we can regulate, the sound (tuning also) is generated by a computer... if it is out of tune (we are talking about stretching tuning, not general tune) means that we decided to have it out of tune.
Some useful links about stretch tuning:
Rhodes piano: https://www.fenderrhodes.com/org/manual/ch5.html
Wurlitzer piano: https://ep-forum.com/smf/index.php?topic=8103.0
CP70 piano: https://my.ptg.org/discussion/yamaha-cp70b-tuning-and-concert-prep-advice
Piano tuning: https://en.wikipedia.org/wiki/Piano_tuning
So... verifying the tuning of a digital piano conceived like we are conceiving our products (super-closed to the original material) with a guitar tuner is not really a good idea.
If we are talking about the sampled material, also here there's something to point out: our approach to sampling is peculiar. We sample a specific piano, an interesting instrument for many reason. Of course they are carefully tuned before and during the sampling sessions but, if that specimen has a specific resonance, detail, noise etc etc... the sample will reflect those. At the end, this is what is making the difference between a sterile copy or something that is coming alive under your fingers.