There are methods of breaking a cipher with enough computation and memory, but those are specific to the pseudorandom function that happens to be used But does

There are methods of breaking a cipher with enough computation and memory, but those are specific to the pseudorandom function that happens to be used But doesn't this constitute leakage of the algorithm's internal state in the output? (Albeit, very slow.) and they are not limited to stream ciphers I never said any of the above was limited to stream ciphers.