In chapter 13 (Quasicoherent and coherent sheaves) of Ravi Vakil's wonderful notes, the author starts by discussing vector bundles, supposedly for motivation. Having understood that each locally free sheaf gives rise to a vector bundle and vice versa, I don't understand how vector bundles "help" tell the story at all. The definition of a locally free sheaf seems to me far more natural, while vector bundles feel (to me) opaque and forced.
In what sense do vector bundles motivate locally free sheaves? Is this simply because most people encounter vector bundles in their mathematical life earlier than sheaves?
As I mentioned in this question, quasi-coherent sheaves are an enlargement of locally-free sheaves into an abelian category. Hence there is a homological algebra of quasi-coherent sheaves and in particular of locally-free ones. So perhaps there are some categories of vector bundles which tempt one to do homological algebra, thereby motivating the sheaf POV..
What are some simple examples of categories of vector bundles that tempt you to do homological algebra?