I just learned a few new things... Why should we not make two notes start at the same time? My notes tend to be 100% quantized, and I use the huminization feature (at max) for any delays and the humanized factor.
I couldn't work with maxed out humanizations scalers. It really sounds out of whack that way. The tuning is way off, like a student band, and a maxed out delay scaler just makes the timing of the attacks very unpredictable and sloppy-sounding. That's not the kind of humanisation one should be after, IMO.
I like to offset the starting points of notes and the dynamic curves purposefully - for example when I'm trying to give a passage of short notes a "behind" or "ahead" of the beat type of feel. Sometimes, I want single notes of different instruments to be completely locked-in and "dead on" - on other notes or accents, I might want a little more rubbing. You can't rely on randomized attack delay to achieve that, and I certainly can't let humanization get in the way of how I intend things to sound.
In fact, one of the reasons I tend to use VSL stuff as the backbone of everything I do is the fact that the programming and editing has been done in such a way that the samples do what I want, how I want it, most of the time. Cranking the delay humanisation would kind of defeat the purpose. Not all sampled instruments are that reliable. A lot of them sound gorgeous, but some of them are all over the place, timing- and dynamics-wise. I hate that, and I find it annoying having to work around erratic attack and buildup times of the samples, especially with short notes.
I love the humanization features of VI Pro, but I tend to dial them in moderately, just for some variation of performance.