Having read the sales blurb and listened to the demos of NotePerformer, I'm frankly amazed to find it has any place at all among the workflows and technical accoutrements of VSL library users today.
I'll put it bluntly: the quality of the provided sounds reflect NotePerformer's cheap price tag; overall, the 'best' demos sound to me like performances on a hypothetical slightly updated version of a Mighty Wurlitzer cinema organ. I simply cannot get past the day-and-night difference in quality between VSL collections and NotePerformer's sounds.
I get the main technical and monetary advantages of using a hybrid of sampled and synthesised instrument sounds, as used in NotePerformer: dynamic articulation and expression changes can be ostensibly achieved using way fewer recorded samples. Also, no doubt the Wallander Instruments devs have done something quite clever with their AI-based automatic phrasing. And so I do kind of get why this product still sells today.
Nevertheless I still remember the often ghastly results, a couple of decades ago, of my best efforts to produce decent-sounding orchestral music from the primitive sample sets of those days. NotePerformer just tends to drag me back towards those days, seeming to make a mockery of the good, steady progress that's been made in the world of virtual instruments since then.
Hope you'll forgive me if I just sound like a sound-snob, but that's the truth of how I feel about NotePerformer. In working with virtual instruments I happen to believe the saying "no pain, no gain" still holds true.