Do we know how careful the Chinese experiments were, relative to the NASA ones? Serious question, because yes, the difference *is* curious... but it's not necessarily due to one result being inaccurate. The experiments were not identical. NASA used much lower input energy, and a non-identical apparatus.
The expressing of thrust in terms of input energy (linearly) is weird and questionable for all the reasons you state. If thrust/energy does indeed remain constant regardless of velocity, then yes, that would appear to be usable for free energy. What this means is one of three things: 1) It doesn't, and we misunderstand the mechanism involved (since the latter half of that statement is almost certainly true, the first could be as well); 2) It does, but this doesn't produce free energy because our understanding of the physics there is wrong (unlikely but possible, if there is a thrust then it's caused by something our previous models did not account for and they only appeared to be accurate because of approximations at near-zero levels of this activity); 3) Free energy is possible after all, and everything that says otherwise is itself not entirely accurate).
By the way, you seem to have forgotten (or misunderstood) that there are multiple drive candidates being tested here. The null device producing thrust anyway indicates that the supposed mechanism of the second drive (the Cannae drive, *NOT* the EmDrive) is wrong. However, according to the inventor of the EmDrive, the Cannae Drive (with or without the slotting distinguishing the experimental and null devices) is basically an inefficient EmDrive. If the Cannae Drive does, in fact, produce thrust for the same reason that the EmDrive does (this assumes, as the experiment supports, that both drives produce thrust) then the supposedly-null device doesn't (dis)prove anything at all and needs no further explanation. Note that this doesn't require that the theories behind the EmDrive be correct, merely that they be less incorrect than the ones behind the Cannae Drive.
Don't get me wrong, I'm as skeptical as the physics as you are... but at the end of the day, the experimental result is what matters. The experimental result appears to disagree with your theories. Therefore, your theories appear to be wrong, or at least incomplete. What you *should* be doing is proposing modifications of your theories and ways to test their correctness. Proposing explanations for the experimental results that are consistent with the current theories (and ways to account for the discrepancy in future experiments) would also be valid. Saying "Nope, the math doesn't check out so it can't happen and anybody who says otherwise is a crank" is just flat-out bad science. We have an experimental result. The result is closer to what the alternative theory predicts than what your theory predicts. The burden is on you to explain that, if you want to maintain the current theory.