To start with, both categories have traditionally been defined through physical parameters. Tools are generally defined as objects used directly by the hand to act upon and affect change in another material. They are purpose-built to do one thing well, simple and closed in their intent. Machines are somewhat different and more complicated. They are broken down into simple and compound and comprised of things that offer a mechanical advantage (that is to say, they make it easier or more efficient to conduct an operation) to a tool. A pulley or lever or bicycle is a machine. Each of these things is designed with a singular purpose in mind, but can be easily utilizable for other things. While tools offer a clear linkage between user and task, the machine offers an advantage in functionality with the downside of less clarity between task and user. Machines are essentially platforms, devices that sacrifice directness for implicit purpose, products with an underlying functionality or framework that, while initially focused in one particular area, can be ultimately applied to any number of tasks.
It's not a great leap to apply that analogy to product design. Each of the products the authors cite as paragons of the "less is more" credo fall conceptually into the tool category. Mies designed homes with a Bauhaus aesthetic. For him the home was a tool for living and not much else. Lotus makes sports cars, themselves single-minded tools for speed and exhilaration. And when Saint Exupéry, reflecting on the earliest generation of aircraft, precarious tools of wire and cotton used to ferry people and mail from place to place, says "Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away" he's literally talking about removing weight so his plane doesn't fall out of the sky. With fewer gauges and frills, the plane is a tool for keeping him aloft. Something like the Flip camcorder is obviously a tool; as the article says, "The Flip had one function, and one only—to fit in a shirt-pocket ready to be flipped out in a trice for those fleeting video moments. This it did brilliantly." The Flip does one thing well; it is a tool for recording.
When approached conceptually, a netbook lies on the opposite end of the tool/machine continuum. It's a platform for other tasks, open-ended, one could even say directionless. The category as a whole has suffered from a lack of focus: Is it a computer? A note-taking device? Does it need connectivity, wireless access? That lack of clarity around purpose has created a machine that can be used for a number of things, but is master of none. And as the article says, "Today's netbooks are now the size of low-end laptops, with up to 12-inch screens, 160 gigabyte drives, one or two gigabytes of random-access memory, a video camera, and cellular as well as wireless transceivers. As a result, they now weigh well over three pounds and need a padded case to lug them around. The original concept has been lost in a blizzard of feature-creep." The netbook is now just a hobbled laptop.
Too often we're creating machines when we need tools. There are a number of reasons for this, everything from marketing pressure to sell the next new thing to hungry consumers to an obvious lack of focus during the design phase. But I think it's also a byproduct of the sophistication and versatility of the tools we're using. It's no coincidence that the examples cited in the article of tool-like products (ultra-simple video recorder, Bauhaus home, retro sportscar, and antique aircraft) are largely physical or defined by their relative simplicity, and that the machine-like devices (netbooks, operating systems) are digital and far more complicated. It's hard to see the bloat behind an operating system when the space it lives in is invisible and largely elastic.
Like most designers of both digital and physical products, I believe that the things we create should live by the credo of "do one thing and do it well," but that's a lot easier said than done in a largely digital world where it's relatively easy and inexpensive to add features. Mobile phones offer a good example: the 'phone' designation is increasingly a misnomer as devices become more and more of a bloated platform for unrelated tasks that consumers may or may not want.
I often wonder if we're entering an era where our physical world will be defined by simplicity and the virtual by relative complexity. But bloat is bloat no matter where it lives. The underlying lesson in this article seems to be, bloat happens, and that more often than not it's going to happen to platforms.