EmDrive is a possible propulsion system granting no apparent expulsion of propellent. It has been reason of hot debate with strong criticisms from the scientific community championed by Sean Carroll and John Baez. 
Notwithstanding this, NASA went ahead performing in-depth experiments to test this technology. It was announced that, on December this year, a paper will appear in a reputable journal reporting on the last measurement campaign. But to today the paper leaked out on reddit. The results, at a first sight, appear to be striking confirming the effect, with all the possible mundane disturbing causes excluded. You can also find other material there as a small video.
I hope this will go properly evaluated by the scientific community moving to a more serious addressing of this effect.
Update: The links were removed from the subreddit’s moderator. I have copies of these files but I do not mean to publish them in any form.
Update: Here is the published paper.
The Gauge Connection 
[…] enterado gracias a un tuit de Lautaro Vergara @VergaraLautaro, que se enteró gracias a un post de Marco Frasca @MarcoFrasca5. Los enlaces al artículo en reddit han sido borrados por un […]
your behavoir is responsible.
Let us wait them to be published as planned, and the heroic published to benefit from his courage.
Off topic:
I’d love to hear what you have to say about why glueballs, one of the earliest prediction of QCD, have been so elusive experimentally.
The familiar pat explanations that glueballs are hard to see due to mixing with quarkonia doesn’t seem to be holding up. At a minimum, it seems as if glueballs are suppressed relative to naive expectations about their branching fractions in decays, by some unknown mechanism, even if they are not outright non-existent.
For example, “oddballs” (glueballs which can’t mix with quarkonia) ought to be possible to seen in current experiments, but are not, and it appears that most or all scalar and tensor mesons can be explained a “molecules” of vector mesons. See my post & comments at http://dispatchesfromturtleisland.blogspot.com/2016/11/glueballs-still-elusive.html
Experiment seems to increasingly support that hypothesis that glueballs don’t exist or are highly suppressed, but nobody seems to know why. Is there any elegant way that QCD could be modified to have this rifleshot effect of preventing or suppressing glueballs without screwing up the rest of the remarkably successful edifice of QCD?
Indeed, is anyone making BSM proposals for modifications of QCD at all? There are lots of experimental anomalies in QCD compared to the rest of the Standard Model (many probably due to flawed theoretical modeling of approximations of pure QCD), yet the amount of BSM QCD proposals in arXiv seems negligible compared to almost any other part of the Standard Model or General Relativity.
Dear ohwilleke,
Your question is clearly off-topic for the post but surely on-topic for the blog. Some readers could miss both it and my answer. Anyhow, I will try to yield an answer from my take about the matter.
As you may know, if pure Yang-Mills theory is concerned (just glueballs and nothing else in the world), the propagator I have found addresses correctly the values obtained on the lattice for the states of the theory. I addressed the question here and was on 2008 in Montpellier. My recent exact solution of the theory keeps the identical spectrum. If in the world quarks would not exist we will be happy with this and can claim to have found glueballs. Unfortunately, things do not stay this way and a pure Yang-Mills theory is not realized in nature. Yang-Mills theory, as any other known gauge theory, comes always coupled to fermion fields and for a good reason. A Yang-Mills theory appears to behave trivially in a low-energy limit. Interaction with quarks makes it not trivial as also happens to the Higgs field (siblings as they are). This fact is not yet accepted by the community even if all the evidences take to this. Most colleagues just claim that the definition of the running coupling in the low-energy limit is not an assessed matter and can hide a non-trivial infrared fixed point just in a new definition. As recently shown by Deur (see here), the fact that the theory shows a trivial propagator but is confining yet is due to the running coupling. Deur just shows that a scalar field theory is able to recover the Cornell potential obtained in lattice computations in this way! So, the idea of mapping between scalar fields and Yang-Mills theory is correct as already proved elsewhere but this evidence is definitely striking.
So, what happens when quarks are introduced in this scenario? The low-energy limit of QCD turns out to be a Nambu-Jona-Lasinio (NJL) model. In a very low-energy limit, just the lowest excitation of the Yang-Mills field enters and produces the
meson currently observed in accelerator facilities and marked as f0(500) on PDG. This is a glueball but NJL says us that it mixes up with quarks. This fixes correctly the mass (see here). Going to higher energies, also the higher excited states computed in a pure Yang-Mills theory come into play and, yes, again they will mix up inexorably with the quark fields. But now we can be able to identify them because these computations can be accomplished. I have to confess that I have not done them and I am not particularly interested in doing this.
About NJL I would like to point out an interesting argument. This theory is known to be non-confining. But if you check the states of the theory they are all bounded states. The question is, yes the theory is not confining, but what if it generates just bounded states of quarks in its spectrum?
Ciao,
Marco
Excellent insights as aways, and it is always pleasure to see Deur’s good work (he has made notable contributions to both quantum gravity and in his main field of QCD). Your insight about bound states in NJL is really intriguing.