You ask what problems Hopfield model has? The following is poorly articulated view.
*Time is crucial for memory and we do not understand time in standard physics framework. Hopfield model accepts the identification of subjective time and geometric time. I have mentioned Libet’s findings many times: they are taken as proof that free will is illusion.
*The identification of all memories as behavioral patterns is wrong- episodal memories are the genuine memories and probably very little to do with the formation of association.
*Synaptic contacts develop all the time. How to avoid the change for longest term memories? It is difficult to understand the fact that the episodal memories of youth seem to be the most stable one. My Grandma literally lived in her youth for several years!
*The neurons in hippocampus (at least) and therefore also their synaptic contents are regenerated. How the memories can survive in this process? I remember also a document program telling about a person had lost almost all his brain and was able to do mathematics!
To me, memory recall looks different from memory storage: in Hopfield model this is not the case since essentially learned behaviours stimulated by inputs as association sequences are in question.