Basically, EM drive would be a very unique method to generate thrust as it doesn't require reaction mass, a feature that can only be found in photonic laser thruster / solar sail solutions, except EM drive wouldn't need any external support devices to move it along, which is the case with huge laser arrays/mirrors/sails needed to propel with PLT.
The paper you linked is being currently disputed and has been subject to many criticisms from peer reviewers before it was finally published. There are several dubious things about the experiment itself, since it hasn't been replicated anywhere else and all teams that did similar research retracted their findings or went silent. Even considering the quality of their experimental device, there's much left to be improved.
As things stand, it appears to be that while there's a potential for the claims about the reactionless drive generating thrust to be true, the theoretical limits of the amount of thrust it is capable of delivering over time make it impractical in competition with more efficient solutions that are in demand due to the nature of contemporary space exploration.
For comparison, assuming the 1.2 * 10^-3 N/KW figure is verified, that makes the EM engine some 33 times less efficient than the most generic of fully electric Ion Engines
(with reaction mass). Since it takes 834 KW to produce a N of thrust and it's 25KW/N for Ion. That puts any potential applications far beyond the scope of interplanetary transfers and one has to consider how huge of a reactor / solar array (power source) it would require to meet any practical requirements.
Here's a nice breakdown of the issue in the spoiler below: