I would add

(c) To convince the reader that I had a good reason for caring that A is true.

This has seemed sadly lacking in many of the papers I've refereed (sample bias, probably). Of course this goal is not easy to achieve, but I do sometimes wish authors would make more effort.