Design is not directly detectable in an object. If I'm out camping and pick up a rock to pound in my tent stakes, I have "designed" a hammer. After I move on, you would be hard-pressed to find out which rock I had used. Even if I shape the rock for the purpose by banging it against another rock you might have a hard time picking it out--ask any paleontologist who is trying to find stone tools in a rockpile. In fact, what he is looking for are traces of human manufacture from which he may infer human design, and when he finds them he may still not be sure of the purpose of the object, what is was designed for.
Considering the pump of your example, I would infer a human designer not because of its functionality or its complexity but because it was obviously a product of human manufacture. If I could not conclude that the object was of human manufacture, then I could draw no inference one way or another about the existence of a designer.
The suggestion that I would infer the existence of a designer in the case of the man-made pump because of its functionality or complexity and deny it in a natural object of equal functionality or complexity out of ignorance or a desire to deny your pet theory is slanderous and offensive.