Slashdot videos: Now with more Slashdot!
The truth is much more complicated. There were literally dozens of variations of shuttle design, and most (even from the very very beginning of the design program, way back in '67 or so) involved the final orbiter hanging off the side of something. Initially, it was because it is hard to design an airplane which doesn't have an aerodynamic nose (the original booster designs were generally airplane-like). Later, it was because the needed tankage for the hydrogen and oxygen to be used by the orbiter was very bulky and would be too difficult to house in the orbiter itself, so they decided to make it external. Obviously, it's hard to fire your engines sitting on top of something, so that meant it had to be on the side.
This, by the way, means that tmosley is wrong. There were always good engineering reasons to put the Orbiter on the side of the ET, which is why that design was adopted. And if he thinks that private industry would have less reason to cost-cut (especially if there probably wouldn't be much risk until you had already made it golden)...
Actually, gigantic chunks of debris could survive reentry--it's actually surprisingly bad at destroying things completely. Look up how much survived on Skylab's reentry, for instance, or how much survived of Columbia.
However, there's no real risk of some dictator flying up and occupying the place, since (given that the Russian fleet is currently grounded) there is exactly one other country currently capable of flying to the ISS: China. And there's obviously no reason for them to use the ISS as a colony drop weapon, since they possess a number of cheaper and more effective nuclear weapons, and in any event would mostly just lose from a major war (imagine both the Russian and American stockpiles being targeted at them...ouch).
There are a few reasons for that. First, the entire Shuttle paradigm--a big vehicle that carries humans and a pretty good amount of cargo, and is reusable--has been pretty well discredited by Shuttle's poor performance. Carrying cargo and humans together has been shown to be inefficient, since the safety standards for each are so different, and cargo just doesn't need people around to manage it, high-profile failures like Skylab, Hubble, or Palapa 2/Westar 6 notwithstanding. People would be (rightly) very skeptical that a follow-on vehicle with the same design could perform much better, even with better technology. That means that any follow-on would probably be much smaller, just a crew transport, IOW similar to the vehicles proposed during the Orbital Space Plane project back before Columbia, or Orion and the other proposed crew transport vehicles.
Second, many of the details of the Shuttle design have proven unsafe. Any follow-on vehicle, for example, would have to be a series-staged vehicle, like most rockets, as opposed to the Shuttle's parallel-staged design in order to avoid damage from foam shedding, as with the Columbia accident, or booster failure, as with the Challenger disaster. Along with the above point, this means that any new vehicle would basically be a completely new design, rather than just a copy of the Shuttle aeroframe.
Third, we've had advances in materials science and aerospace engineering that mean we could do better now in terms of the details of the Shuttle's design that they could back then, many of them gained due to Shuttle experience. We've flown a winged vehicle through high-Mach regimes at very high altitudes in the Shuttle program, something that hadn't been done before. So, by using a new design, we could produce a vehicle that did better than Shuttle. Again, a reason to simply not copy Shuttle with better internals.
Fourth, doing so would be very expensive. Since, as noted above, the Shuttles have not been particularly successful, there's no reason to spend a lot of money copying them. Instead, people are spending money on copying Shuttle's big unique capability--ie., crew transport--while cutting out all the irrelevancies that cost a lot of money. Even then it's expensive, but you skip the need to design a lot of stuff and it works out to be cheaper than trying to also build the carrier rocket, a big payload bay, etc.