That's one of them, yep.
That is one of the factors excluding theism as a religion - too many theistic religions, all different.
All right: so your position is that there are too many theisms for atheism to be called any one thing; that atheism cannot then be narrowly defined.
But - related to the above - any given thing can indeed have a multivariate definition. Cannot "atheism" then simply be the origin on multiple axes of belief, as proposed by Dawkins for a single axis of belief? I don't think it can be excluded from the process because it doesn't correspond to a single description.