Case studies are by far the easiest kind of history to do, and do well. There are fundamental difficulties with writing surveys, largely because it is so easy to make generalizations from insufficient data - and even if one is swimming in data, how that information is read depends entirely upon how it was gathered and the way it was processed, as any statistician could tell you. It's possible to achieve a decent balance between the extremes. The social constructivist school strives to do just that, for example. In the history of printmaking, Adrian Johns' The Nature of the Book does a decent job of that balance.
At the same time, however, I do believe that survey texts are necessary, especially as a foundation for a field. We don't all have time to read a few dozen carefully selected primary sources to give us a sense for how a period operates... although this is certainly an admirable way to work for an field or time period where getting to know the worldview and politics is important. Surveys are brave and admirable undertakings, always open to attack for all the relevant literature the author failed to read along the way, since the large the subject, the greater the challenge of doing all its components justice.
Even in survey works, however, those which venture outside the confines of a country or region are the rare ones. David Nye's book, for example, may be a survey work, but it sticks to discussing America. Adrian Johns stays with England. A history of the way a given technology has affected the entire planet requires an understanding of both the technology and all the cultures involved... and this planet has a whole lot of cultures to offer (to say the least). Consequently, many of the popular large-scale survey works are either dismissed by a large smattering of interested historians or else used judiciously, with large chucks of salt. (I'm thinking of Jared Diamond's Guns, Germs, and Steel: The Fates of Human Societies when I write this in particular, one of the more reputable among the contentious planet-wide history pieces.)
The challenge of survey work has been on my mind frequently as of late, a shadow off on the horizon with which I must grapple in the next year since my dissertation topic is a survey piece, spanning several hundred years and half-a-continent. I only hope I can do it justice.
Note: I've always been mildly leery of claims for the "most important technological inventions ever." They're handy as a teaching aid, but rarely do the complexities of history justice. Most of the items or processes which end up on the "most important" lists are those which are easiest to see in daily operation: books and time-keeping methods are perennial favorites. Electricity is also a fairly popular one. The telephone and steel-making aren't up there quite as often, but they often figure. So do semi-conductors. Off the top of my head, spinning, weaving, writing, boats, the domestication of animals, coinage, sewage systems, and indoor plumbing are some of the many more neglected major ones.