It's tough to win someone over in the God debate, and I don’t really ever expect to, but I do think it should be possible to make progress on the appropriateness of the terms agnostic and atheist. Apparently this is too much to hope for and people like Jane Galt irritatingly call themselves "agnotheists" and stuff like that.
Anyway, here’s another attempt. At the end of the day most people assign a rough probability to the existence of God. The mysterious nature of God doesn’t change the fact that P(God) is between 0 and 1. So it should be possible to compare your P(god) with some other everyday probability and then choose the everyday probability you consider closest to your P(God) and then use similar language about both probabilities.
For example, P(heads) = 0.5 and I am “agnostic” about the result of the next toss. P(lottery win) is very low and easy to calculate if I buy one ticket. I am not “agnostic” about winning the lottery, I believe I won’t. If I say, “I believe I won’t win the lottery” people won’t jump down my throat because I MIGHT win the lottery; I know I might!
Jane Galt thinks that P(God) is vanishingly small (which is exactly the same as the atheists she finds so objectionable) yet she can’t bring herself to say, “I believe God doesn’t exist” because somehow that would imply that she thought she could prove that P(God) = 0.
Shouldn't she should know that Dennett and Dawkins don’t think they can disprove the existence of God. If she were Christian would she refuse to call herself one because there are some dogmatic, unpleasant Christians?