You are 100% correct. It's an easy trap to fall into, to only verify the subspace conditions instead of all the many conditions for a full vector space, but just verifying subspace conditions is definitely not enough! I've seen countless students (as well as a few teachers) fall into this trap.
Now, there is one potential saving grace here: perhaps it was proven that the set $W^V$ of functions from $V$ to $W$ (linear or not) is a vector space. That is, $W^V$ under the given operations, all of the 8 or so properties of vector spaces were proven previously. Then, proving $\mathcal{L}(V, W)$ is a subspace $W^V$ is a valid way of proving $\mathcal{L}(V, W)$ is a vector space in its own right, as it will inherit almost all the properties from $W^V$.
But otherwise, the proof is incorrect.