In fact, exactly such an effect was experimentally verified in an 1895 study of Axel Oehrn*. While subsequent experiments have confirmed this result, it depends sensitively on the type of material being memorized, on the exact time intervals, and many other variables. Now, in some sense this contradicts the Ebbinghaus exponential forgetting curve. In practice, a pretty good heuristic is that the Ebbinghaus curve holds approximately, but there are exceptions, usually over limited times, and for very specific types of materials. I don't mention this to undermine your belief in the Ebbinghaus model. But rather as a caution: memory is complicated, we don't understand many of the big picture questions well, and we should be careful before we put too much faith in any given model.