The distinction I'd elucidate from the gp's is that "great" is a problem comes because it's both a superlative and because it can't be quantified.

You can ask for a developer who produce a large amount of average code. You can ask for a developer that produces very few terrible errors. You can ask for a develop who produces decent code and happens know domain X very well. You can ask for a developer who can mentor younger developers. You can ask for a developer who might needs virtually no supervision to succeed. And so-forth. They could all be different people. Perhaps you can sift until you somehow find someone who's all that but they'd be fool to join you because you're looking for perfection rather than looking for fulfill your actual needs.

Superlatives are a terrible way to make any kind of serious decision. Salespeople love specifically because they tend to be unmeasurable and irrelevant when they are measurable.

Thus the more superlatives are entering a decision process, the more you can tell you're sold shine rather than substance.