PaintCode is meant to replace the need for images by giving you the equivalent code for basic images. Their initial claim to fame with version 1 was being "Retina ready," since the images are drawn in code instead of PNGs. It is a very good way to learn Quartz and CoreGraphics, but not necessarily the best way to learn ObjC.

I used PaintCode in a project I worked once, and it was a lot more work than I wanted for just asking my designer to give me an @2x image. If you work with designers who do vector art already, you're not going to gain much. The StyleKit class is slick, though, so I may revisit this once again.

This was our experience as well. We used it for a few things, but in the end it was faster for us to just create a @2x image.

I am wondering if PaintCode would allow my designer to draw up a complete UI - e.g. all screens so I could just use that.

What I do now is I get the designs in some sort of intermediary format, e.g. InVision or whatever, then I have to go and measure everything and all the pixel distances, then re-create it in Interface Builder. It's retarded.

Couldn't you just use Interface Builder to begin with? You can e-mail or put into git .xib files right?

The greatest benefit I can see with PaintCode is the ability to make parameterise-able images with expressions. Creating cool little dynamic vector images that can be driven through your code (gestures, accelerometer, etc) seems really nice.