Not necessarily, of course. But we very often see scientists grossly underestimating the complexity of the unknown, the stuff left out of their model - same as anyone else. Scientists have no advantage over anyone else in this regard - even their presumably higher than average intelligence is often of little help, as it often acts to provide false confidence and prevent recognition of the situation they are in.
Examples are legion. Here's one:
https://dspace.mit.edu/handle/1721.1/6125 The description might now be assumed to be describing a full scale research program for years to come, but at the time it actually meant what it said: visual object recognition and naming by a robot equipped with a camera and hooked up to a 1966 computer was assigned to a few students ("summer help") at MIT as a one summer's project - software, hardware, the whole shot.
We have hints of a somewhat different, complicating factor
http://galileo.phys.virginia.edu/~rec_ak/Lab_Safety_Site/rubberband.jpg
that may have been very important - in many cultures, early stages of what we now recognize to be entire fields of technological advance exist only as toys, playthings and/or recreational gizmos, maybe musical instruments or decorative curios or the like. The Inca invented the wheel and axle combination, for example, but had no known adult use for it - they made wheeled toys for children. The Chinese invented gunpowder, and used it for fireworks. Would anyone be surprised to discover that the forerunners of the axlatl and similar devices elsewhere were sticks used by children to throw mudballs and crabapples and the like at each other?
The way it seems to work is: some people like to fool with stuff, and they figure things out that later - when necessity comes along - suggest possibilities that otherwise would not have occurred to anyone. That suggestive effect depends on a degree of abstraction - some development of what amounts to theory.