In the derivation it is assumed that all the rays meeting at a point are parallel but then parallel rays don't meet at a point and no interference can take place if they don't meet at a point.
So if they meet at a point then they are not really parallel if they aren't parallel that means that subsequent rays coming to the same point would nat have the same path difference of exactly lambda thus no bright spots would be seen.
So its crucial for us to assume the rays to be parallel but in reality its not possible and even a small difference can not be neglected here as a lot of rays are interfering at the same point so the rays need to be exactly parallel so How is it justified to assume that the rays are parallel in the derivation?