The supposed problem of the is-ought dichotomy is that statements about reality (what is) are fundamentally different from statements about morality (what one ought to do). A supporter of this position would go on to say that we can't figure out anything about morality on our own because we can't deduce an "ought" from an "is."
For example, "mercury is poisonous" would be an "is" statement. But, this view would ask, what does that have to do with morality? Should we drink it? Should we not? A supporter of the is-ought dichotomy would say that no matter how long you stare at the three words "mercury is poisonous," no "ought" will appear—you can't deduce anything that didn't already exist in your premises—so you have no way to conclude whether you ought or ought not to drink it.
Rand dealt with the problem by attacking it at the root. According to Rand, the bridge between the "is" and the "ought," between reality and morality, between fact and value is the concept "life." If you choose to live, that instantly implies a whole slew of values, i.e. "oughts." One of those "oughts" would be that you shouldn't ingest poisons, mercury included, because if you do, your life will soon go out of existence.
Coming back to the topic of this question—Rand's statement that every is implies an ought—she really does mean "every." Even boring, trivial facts such as "the sky is blue" imply "oughts." What does "the sky is blue imply"? One implication would be: if look outside during the day and the sky is dark, that means that you ought to take your umbrella if you go out. (Leonard Peikoff discusses this point in the Q&A of his Moral Virtue lectures, from which I take this example.)