I don't think it's redundant.
If you say, e.g., "It is likely that Sally Jones will win the contest", you are saying that you are not certain that she will win, but you think that the probability that she will win is high. But if you say, "It appears likely that Sally Jones will win the contest", now you are saying something more like, Based on the information available at present, the probability that she will win is high, but we are not certain of that information. Or perhaps you want to allow for the possibility that between now and when the contest is actually decided, circumstances may change. Or you simply want to hedge on your estimate of the probability.
It is true that any statement of probability -- "it is likely", "it could happen that", " maybe", etc -- implies that the event is not a certainly. That is, if you say it might happen, the fact that you are not being definitive means that you are also acknowledging that it might not happen. In that sense, adding additional qualifiers could be said to be redundant. Like, how is "it is possible that" different from "it might be possible that"? Etc.
But different wordings imply different probabilities. Of course there is no exact formula: we do not say that "it is likely to happen" means a probability of 80%, "it may happen" means 40%, "it might happen" means 34.7%, etc. But "likely" implies a fairly high probability, while "appears likely" would mean something less. "It might happen" is (probably) more likly than "It might possibly happen", which is more likely still than "it is barely possible that". And so on. We combine various words to shade our estimate of the probability.
I am reminded of the time that in a conversation someone made a suggestion, and I said, "That may be an idea". Someone else laughed and noted that I wasn't even willing to say that is was an idea, I would only concede that it "may be" an idea.