the former assumes an absence of questions. according to the merriam webster online dictionary (oh, for the days when people turned the pages of an actual dictionary!), certainty is "a state of being free from doubt," and certitude tops a list of synonyms (more on that later). doubt, on the other hand, leaves open possibility. in a state of doubt, there are options: this or that may be. among other things, m.w. defines doubt as "a deliberate suspension of judgment." and a suspension of judgment further implies the possibility of un-suspension; that is, it is a temporary state. judgment is possible but is being withheld for the time being.
of course, in most any circumstances, certainty is preferable. i'd certainly rather be certain about my plans for next week or about what i'm making for dinner or about what time my upcoming meeting will end. unless, of course, those things about which i'm certain are not the certainties i'd prefer. if my plans for next week certainly include spending several days sick in bed, i'd prefer to live with some doubt in the interim. if the dinner i'm making will certainly turn out badly, i'd rather not know for sure before i start cooking. and if my meeting will certainly last several hours, well, i'll choose to remain in the dark until then.
but to go to said meeting, to perpetuate the example for a moment, uncertain of the length (as i might say to sam: "i doubt it will last too long") leaves room for possibilities: it is possible that this meeting will last several hours, but it is also possible that it will move quickly and efficiently. though i might have little hope of the latter and even perhaps expect the former, i can retain some hope for a brief meeting as long as i'm not certain of the outcome. so, although i'd rather not live with doubt, in this case, i'll choose to cling to doubt as a window into possibility.
such is the case with questions of belief, i think. certainty--or perhaps certitude, which, as m.w. suggests, connotes "a faith in something not needing or not capable of proof"--is preferable for sure, as long as the fact of which you are certain is the result you would prefer. to return to my simple example, if i can be certain that my meeting will last only half an hour, then i'll take certainty over doubt for sure. but if, on the other hand, the only certainty i can imagine is that it will last three hours, then i'll choose to retain my doubt--and thus my hope--as to how long it will or will not last. as for more important questions of belief, then, if the only certainty i can imagine at a specific point in time, for whatever reason, is not a pleasant choice or the result that i would prefer, better then to choose doubt and its accompanying hope that the unpleasant certainty i can envision is in fact not the only possibility.