It seems like you already can see this, but Ext^1(M,N)$Ext^1(M,N)$ is measuring all the ways to form distinct short exact sequences 0 -> N -> ? -> M -> 0$0\to N\to ?\to M\to 0$
If you are looking for intuition, what you should do is think about picking a set of generators G$G$ for M$M$. This is equivalent to choosing a surjection from the free module R^G -> M$R^G \to M$, whose kernel is the 1st syzygy module S (which you should think of as relations between the generators). Then Ext^1(M,N)$Ext^1(M,N)$ is going to be all the maps from S$S$ to N$N$, modulo those which come from a map R^G to N$R^G \to N$. Therefore, Ext^1$Ext^1$ is something like all the ways of assigning an element in N$N$ to every relation between generators in M$M$, modulo dumb ways of doing this assignment (those coming from sending generators to N$N$). This perspective can make it easy to guess what Ext^1$Ext^1$ should be.