This debate is covered in the discussion around the original articles. And as I have mentioned numerous times in this thread, there is an argument to be made for the double amputee having a HIGHER cost of running than an able-bodied athlete. So if you go back on the thread and read the post, you'll see that there are effectively two models - one says higher energy cost (your argument, explained above), the other suggests lower energy cost.
Now, the data that the IAAF have collected can be used to evaluate the models - quite clearly (and correct me if I'm wrong), if the cost of running is HIGHER, then he would have to slow down more than the able-bodied runner, because the metabolic cost is, after all, the thing that causes him to slow down in the first place. That would be like comparing an efficient athlete to an inefficient one - the difference is manifest as the time of exercise increases.
However, the energy cost is LOWER, then the model predicts that he would be able to sustain the same speed. And that's exactly what happens. So I hear your argument, it needed to be considered, but it's not been confirmed or even suggested by the observation. And if you really think about it, to suggest that the energy cost is 25% higher, you are saying that Pistorius is actually 25% better than the able-bodied runners he wants to race against. Even if only 25% of the performance could be accounted for by energetics (which is an underestimate), then you are saying that Pistorius should be capable of running 6% faster, which is 12 seconds - so that means we are watching a 36 second 400m runner! That's logic I'm sure will be questioned, but the numbers might be questioned, but the logic is there.
The other thing you are perhaps missing is that all these studies have been done examining non-running prosthetics. There are massive differences between these non specific prosthetics and the running ones, like the Cheetahs. So the very latest work, which was presented at the ACSM conference in June, showed that the running prosthetics are able to reduce the difference in economy between able-bodied runners and amputees - the difference disappeared at 8 miles/hour pace. Now as the speed increases, the Cheetahs will provide even more recoil and return of elastic energy. That's how they've been designed. So all this means is that there ia ctually scientific evidence that the RUNNING prosthetics, which are specially designed to propel the athlete, store energy, create extra bounce, built in shock absorption etc. are actually better, even from a data point of view.
So your studies are correct, but you have to realise that they are different, and that difference, is, to be redundant, the difference!
One final thing is that anyone who loses a limb/s after learning to walk will probably never be able to learn to sprint. Again, this has been said on the forum already, but it's worth repreating - the motor control of the limbs is probably too complex to RELEARN, and so for anyone who has to learn to walk AGAIN, the complexity is what causes the abnormal gait. Now, if you learn to walk on prosthetics from a young age (try one year), AND you have a physical therapist to train the 'bad habits' out of you, then you are in a postion where all these differences are minimized, if not removed completely.
So the point is that a lot of research that you cite and have linked to is correct, but this is a competely different scenario - you just have to look at the pacing of that 400m race to realise that.