1. Scatter behavior changes on zoom.
Maintaining the same parameters and behavior scheme, I have seen a few instances where the pattern drawn will change distinctly at different zoom levels. One example is stemming from the 'from spherical' cast operator, shown below:
- ex1.png (219.9 KiB) Viewed 10649 times
Zoom levels were set to 25% increments.
I've tested it with another helical brush like that one with similar results. It seems that at different zoom levels the helix number is changing (more or less spirals at different zoom levels). I'll post the test brush for reference.
[Edit: Looks like this is related to 3 and 4; since the multiplier is so high the frequency is going crazy.]
2. When manually adjusting the zoom my cursor kept redirecting elsewhere after clicking in the percentage box.
I think what's going on is: I was using a mouse to click inside. My mouse sits on a smooth surface so it moves slightly when I take my hand off it. That small movement was causing it to deselect the number a moment later, and the tooltip would pop up. Then I would type the number in and it would start selecting brushes. This is all happening in just a couple of seconds. Also, when I would hit enter after typing in the number it would change the zoom level but not turn off the zoom function. So then without realizing I would try to draw and end up readjusting the zoom.
3. Slider resolution is higher than number displayed.
Ok, so this is a tough one because it hits at another question. I sympathize that at some level all rendered lines are a series of points unless expressed in vector form, so shape interval will be extra sensitive. At that, what is the true spacing interval when the slider is set to 0? Do all points still follow the behavior scheme as defined at that resolution? For other things like size, opacity, and scale I'm seeing that I can very, very slightly adjust the true output without actually changing the number shown on the slider.
4. Sin
I'm thinking this is related to 3. When changing the period and watching the drawn function in the sinusoidal node, it goes way crazy when approaching 0 (to be expected). But feeding length into the sin function gets really tricky, since that resolution issue pops up as irregularities in the stroke.
5. Controller menu
First, wow look at all the new options.
Second, wow look at all the new options. 
Haha, it has expanded so much it's becoming tedious (as a controller troll) to scroll through the different options. Can we get an option that will just show it as a simple list sans icons? Also, (and I know you all will be explaining what the new nodes do at some point, so this might be jumping the gun a little) it seems that all the yellow coded nodes are scatter related... Can we have an option to separate these from the orange (orange menu, yellow menu)?6. Brushes parameter change having no effect!
Someone has posted a similar problem in this forum, and I didn't realize that this is something I had been seeing all along, too. Size is a good example. Obviously, you shouldn't be able to set the maximum size -smaller- than the minimum size. This works great. But after trying to force it (intentionally or unintentionally), sometimes it will stick. It either won't accept input, or the number will change but the output won't. I noticed, though, that if you open the controller panel and stick any node on the problem parameter then remove it, the parameter will no longer be stuck. This was actually a -huge- source of confusion when trying to learn Bi in the beginning. It might not all be related to the min/max sliders, though.
That said, it's all very exciting. Loving it.
this is a well know problem and will be fixed 
So the stroke modifier will be vector based and not have the high frequency issue?
This is what I gathered... What we're seeing are the aliases (harmonic reconstructions) of an undersampled (low resolution) signal (brush function). Increasing the resolution/dpi would increase the sampling and would (hopefully) allow the nyquist criterion to be met. That's why I saw multiple iterations of the same pattern when close to 0% zoom on the FromSpherical scatter brush, and zooming in eventually caused it to be undersampled (same with the candy cane). Thanks again for the trouble. It's nice to have words to put to the idea.