Yes, I've seen the video before.
The problem is EVERY explanation yadda-yaddas how you collapse the "vector" or array. The algorithm can't define a probability state that's why it's called a probability not a result. The algorithm he uses in the video also gets "closer" to a result state but never arrives at it. Which is useless in computing. This is why I asked the question and much like the OP, I don't want to watch another quantum computing video. Just answer the question.
Don't mind me If I don't expand too much.
The usefulness is due to other gates such as the haddamard gate and unitary operators.
In the case of shor's algorithm, it brings back the complexity from exponential to polynomial (O(n^2log(n))), making large prime factorization much faster but not costless.
Okay, if that is true the probability has to collapse to 1. My point is you can approach the probability of 1 but it never is absolute. This is the fundamental problem (yes even using those quantum gates) with a quantum array or "vector" the inability to use the information even if you can approach a closer probability.
A high enough probability is always good enough.
