Practically speaking, you can think of this form of propulsion as a microwave oven with the door open. Right now, running it on "popcorn" with the governors off MAYBE produces tiny amounts of thrust in the direction you point it. The idea is that everything starts off producing tiny amounts of whatever and maybe eventually the amounts won't be tiny. This is potentially useful because the laws of thermodynamics require you to throw mass in the direction you don't want to go in order to move in the direction you do. You can throw tiny amounts of mass super-stinkin' fast (like with ion thrusters) or gob-smackin' amounts of mass at poky speed (like the stuff we tend to do in orbit) but for every action, there is an equal and opposite reaction. The EmDrive appears to violate this, which is beneficial because theoretically, you wouldn't have to haul your reaction mass into space; you could throw out a couple solar panels and head to Pluto on a steady diet of sunshine and good intentions. It would not, however, a hoverboard make. Assume you increase the efficiency a million percent. 850 watts would give you a little over a kilogram of thrust. If you wanted to lift, like, you you'd need to point the equivalent of about 80 microwave ovens at the ground, which would certainly cure any weed problems you may have but also be substantially less convenient overall than essentially every other form of thrust you care to think of.
It has been reproduced by NASA Advanced Propulsion Physics Laboratory, Northwestern Polytechnic University in Xi'an, and Dresden University of Technology.duplicate this experiment in different labs with different equipment.