Let's say I iron my shirt or dress or tablecloth so that it looks great: nice and smooth and smart. And then I wear/use it. And then I suddenly notice that I missed a bit and there's a wrinkle in it. The way you'd feel about that literal wrinkle is the same way you'd feel about a figurative "wrinkle" in your data. It's something that you hoped wouldn't happen, it's an anomaly. It's not major so it doesn't ruin everything. But it's ironic that you tried really hard to make it perfect and then you spot a wrinkle. Business and science uses this figurative "wrinkle" for anything that's minor but spoils what would otherwise be perfect.