Rather than analysing the content of dreams, I think the dreamstate is likely a way for us to improve our learning capability by injecting noise (randomness).
Give this machine learning article a read, I find it very similar.
https://machinelearningmastery.com/train-neural-networks-with-noise-to-reduce-overfitting/