I don't think this is a good heuristic. I would call it a fallacy; I don't know if it has an official name, but in my head I've been calling it "syntactic similarity implies semantic similarity" (this being the fallacy). Ideas that sound similar don't necessarily map to close points in ideaspace.

There's been some LW discussion, e.g. Talking Snakes: A Cautionary Tale, about the specific application of this fallacy to ideas that sound similar to obviously absurd ideas.