Short Answer
To ascribe the outcome of a current situation to a prior one based on a superficial analysis is hasty generalization.
Long Answer
To expound, a hasty generalization happens when one creates a rule based on an insufficient sample. With a substantial sample, one can use inductive reason to make a probabilistic claim. A man comes home and parks on the right side of a wide driveway 24 days in a row, it's not unreasonable to venture a claim that he will do so on the 25th day, particularly if other regularities and inferences exist. But, to venture that same claim after the first time he parks in such a fashion would be a hasty generalization. Let's take a look at your example.
When someone believes something is impossible, someone else claims that "people believed flying was impossible until it became possible too so" as though anything that's believed to be impossible is "possible"
So, is it true that some people somewhere claim that flight by machine was impossible, and then did the Wright brothers accomplish flight about 5 years into the 20th century showing such a claim was wrong? Absolutely! This is ONE instance where that which was claimed impossible was accomplished. In July of 1969, Apollo 11 landed humans on the moon again demonstrating what SEEMS impossible is often possible with ingenuity and adequate resources. But, does that mean based on these two prominent examples that all claims of impossibility are inherently flawed? Of course not! That would be a silly conclusion, because these are just two examples, and to generalize their outcomes to all outcomes would be rather hasty.
Let's say the claim is now "Fusion is impossible, because it's been 20 years away from happening for the last 40 years now." Certainly, one might retort, this isn't the first time what has been claimed is impossible will be shown to be possible. But can we draw the conclusion that fusion will certainly happen? Absolutely not! First, it might be a false analogy to draw parallels between flight and fusion. Second, there might be unknowns about subatomic particles that fusion researchers are unaware of. Lastly, there may be unknown unknowns! Engineering proceeds by prototyping and then failing and correcting often using abductive reasoning to move forward to the fulfill the problem specification. That's because 1. Reason is defeasible meaning that one simple fact can completely defeat an argument regarding a prognostication. 2. Empirical evidence is often discovered fortuitously.
Now, the final important thing to recognize when arguing with informal logic is that matters aren't black and white, but rather are strong and cogent to a degree. One can certainly draw parallels to flight and the moon landing to butress one's argument, but one has to do so very well, by showing through argumentation by analogy why the prior instances lend weight to the conclusion. And there's no royal road to creating an argument to draw a conclusion. It's quite an art form to build a solid argument, and even then most philosophers characterize knowledge as fallibilistic.