The previous version I did of svgslice allowed slicing by drawing rectangular areas to slice, AND slicing by the use of guides.
The guides are quite inflexible, though, and and became largely obsolete when the rectangle-based method was done. Rectangles allow overlapping areas, and specifying things like the slice name in the rectangle's properties. Eventually, it could allow things like setting the output format for that slice, and its compression level, etc.
Guides *are* a bit faster for simple stuff like just slicing an image into two halves. But I'm a little concerned about the confusion for users of having two separate methods, etc.
Would anyone miss the guide-based method if I left it out and focused on the more flexible method? As long as you can snap rectangles to a grid, it shouldn't be too different.