These experiments were apparently well designed and thoroughly done - they just don't apply to actual EM-Drive but speculative [infrared laser version](https://www.reddit.com/r/Physics_AWT/comments/iz2d4x/darpa_l...) of it. There are [good theoretical reasons](https://aetherwavetheory.blogspot.com/2010/11/do-galaxies-sh...) for why EMDrive should work better with radiation longer than wavelenght of CMBR, i.e. vacuum fluctuations. Also the geometry of resonators tested is [very distant](https://i.imgur.com/zTrRFum.gif) from conical shape of EMDrive, mildly speaking - and actually nonsensical with respect to EMDrive theory, which is based on polarization during reflection under Brewster angle. I'd even suspect, that Tajmar published negative results for misleading designs intentionally for to slow-down Chinese competition in this research. Most probably they already tested conical resonators - because there is no good reason why not to simply do it - but they hide their data before public for to get more time for further research of it. The pathoskeptical Popular Mechanics journal repeatedly attempted to [dismiss if not discredit EMDrive technology](https://www.popularmechanics.com/science/a15323/temdrive-con...) and now it just jumped at new opportunity with all vehemence. The path to replication of original [R. Shawyer's EMDrive drive design](https://i.imgur.com/inLGS5G.jpg) is thus still fully opened.