There’s certainly a lot of media attention being paid to the claim that NASA researchers have validated a new type of superefficient thruster (in theory, it needs no propellant at all, just a source of electrical power). I'm very, very cautious.NASA hasn't officially put a stamp on this: One NASA team says they see a very small but consistent effect. As for me, I want to see it replicated independently. Even if that works, then I want evidence it can be scaled up to a useful thrust level. So it’s very interesting, but this minuscule effect (if it’s proven to be real at all) isn’t ready for flight time.
Here's another thought. Assume the effect is real and can be scaled up enough to put a small thruster on a nanosatellites, like one of the popular CubeSat-based models. If you can deploy enough solar cells, you can use that on inner Solar System missions without violating the laws of thermodynamics, since you are putting in more power than you take out (Given the size of the effect as measured by NASA, there's clearly a major loss of energy in the system). For larger thrusters, though, or for missions beyond Earth, you'd have to carry a power source for the electrical energy needed. Does the mass of the power source (RTG, fission, whatever) cancel out the gain of using this effect for propulsion? It might still work, especially on science or cargo missions where it's ok to take a long time to accelerate. We're still around Technology Readiness Level (TRL) 2 here on NASAs scale of 1-9 (see the diagram below). Arguably, this idea still has one foot in Technology Readiness Level 1, since the basic research needed to understand the nature of the propulsive phenomena is far from mature.
With very tiny effects, even highly qualified researchers can see what they want to see. (This is precisely what happened with the Podkletnov antigravity disk, which NASA put at least $650K into before figuring out the effect was imaginary. Some smart people at NASA believe there's something to Andrea Rossi's e-cat cold fusion device, but it slips year after year without producing the kind of demo that skeptics can't refute as error and/or fraud. )I'd put a modest amount of money into this to see what an unrelated group of engineers could do about replicating it. Replication is always Step One.