Note spelling in scores drives me crazy – that is: seeing how students notate accidentals, choosing between enharmonics. Most frustrating is the excuse, “that’s just how the computer did it.”
At the above site is a link to an interesting, common-sense article by Peter Thomsen outlining a good approach to selecting which enharmonic notes to write for ease of sight-reading. It’s probably most useful for non-triadic music.
Another approach that’s been outlined from time to time is to keep to using the same accidentals for a passage of music, that is, sharps for an entire section. Personally I dislike this method as it can result in ugly intervals such as diminished thirds (e.g. A#-C).