Oh forgot to say before, that’s a pretty standard way to model eyes. I call it “the bowl technique” Did you figure this out by yourself?
That’s neat!
My variation of this technique is to use two concentric balls. One fully spherical, with glass shader with tweaked IOR and no shadow, and 0.001 bigger than the one with the “bowl”
Works for realism and cartoonism alike, hope you find it useful 
Thanks again all for the tips, I’ll surely try to make use of them next time.
The reason I stopped updating this thread is that I kind of moved on in the sense that I’m trying to improve the modeling process by creating new models, so unfortunately I don’t think this one will really be “completed”. For example, instead of per-vertex modeling which I used for the head I showed in this thread, I recently started exploring sculpting, which I used for this one (as you can see from the very high vertex count - no retopologization attempted yet):
But for a different model, a different thread would be better and… as the above model has reached a satisfactory grade of completion for what could be considered an excercise, I’m again working on something else at this moment. Perhaps I should open a sketchbook thread?
I do use two concentric spheres too, but I haven’t thought of using no shadows for the one with the glass shader. I will try that since the outer glass layer tends making the inside too dark and noisy, which has to be compensated for.
I initially attempted to imitate real eye anatomy, but then eventually realized that it’s often done in 3d modeling. In a way, it’s also how BJD eyes are actually made: they’re not (at least usually) simply flat planes, they have a 3d structure behind the glossy layer.
I am glad to help
By the way, lemme just tell you that to disable individual objects’ shadow go to properties editor-object-cycles settings.
And you should definitely make a sketchbook, it’s good to get feedback not having to finish anything nor create new threads, and learning how others evolve can be so inspiring! And it’s specially good to remember old works, so we can “revive” them later 
See ya
If the scalp is UV’d you can paint maps just to control the hair color. It’s also possible through nodes to do differences in hair color from root to tip. There’s also ways to map other hair qualities, typically it’s controlled by a grayscale. Might offer more control in some cases than doing it the other way. Maybe you know that already, but I’d thought I’d point that out if you didn’t.
Other than that, it’s alright. I’d say to toggle on hair smoothing or increase the resolution before going to render.
I knew about the hair info node, I often use a color ramp with it and the intercept output, although mostly in a subtle way.
Other than that, it’s alright. I’d say to toggle on hair smoothing or increase the resolution before going to render.
What’s hair smoothing? Do you mean B-Spline?
EDIT
By the way, I’ve posted my latest, hopefully finished piece here: