The majority of my captures tend to be regions.
When capturing regions, the current small cross-hair makes encompassing irregularly-shaped objects difficult.
A nice feature of some screen capture programs (e.g., Hypersnap) in region capture mode is vertical and horizontal guidelines that extend from the edges of the desktop through the center of the crosshair at the start of the capture. That way, the user can easily see whether the capture will include the desired areas.
I have begun to use the RedBox capture mode, but the above 'guidelines + crosshair' functionality in Screenshot Captor would streamline the region capture process quite a bit.