i think a dark side to "finding hay in a haystack" is the risk of building up an elegant theory for e.g. a novel calculus of hay, only to realise after some time that the set of hay is in fact empty, or is isomorphic to some other well understood and hence uninteresting structure.