Well, it's great if you want to generate lots of ozone and want almost no thrust
Electrohydrodynamic propulsion more interesting re: thrust when it comes to airships (the thrust is low per unit surface area, but in that case you have a
lot of surface area). But it's always going to be an ozone generator. If I recall correctly, the three-electrode designs are less polluting than the two-electrode designs.
On the upside, EHD propulsion is surprisingly efficient when surface area isn't limiting. You have a lot of mass moving at low velocity rather than a little mass moving at high velocity, and the former leads to more efficient propulsion. Of course, you have to subtract the added losses from the EHD system...
I was last looking into EHD for the application of propulsion for a Venus colony for
a book I was working on. It's promising for that role, but just not mature enough (probably the biggest problem is that electrode longevity is still a big challenge). Interestingly, it might have some potential for simultaneous propulsion + carbon sequestration - decomposing CO2 to CO and atomic oxygen. Carbon monoxide is highly reactive at Venus surface temperatures / pressures and might have some potential for being sequestered into surface minerals to form carbonates (although I'd more expect it to be consumed in reducing surface oxides). It also
might be possible to generate graphite dust as a propulsion waste product, although configs that would encourage that haven't been explored. This is all an area that would need a lot of further research and should be taken with a grain of salt without further studies.
For Earth applications, I'm more sanguine about electric ram-arcjets. Got a CFD sim running right now on my compute node at home.

Unfortunately, it's in an genetic algorithm to optimize the shape, and evolution has a nasty habit of finding ways to cheat. In this case, it discovered a way to make the model omit its end caps, which made it not an "airtight model", and thus left normals ill-defined, which confused snappyHexMesh and thus rhoPimpleFoam. Net result, a powerful current of air was just streaming from the walls straight into the engine :Þ I had to scrap like a month's worth of work and start anew after fixing that bug.