We could build a decent fusion pulse drive today if we had higher temperature more compact superconductors and super-efficient compact lasers that could fit in a spacecraft and ignite a strongly net-positive inertial confinement fusion pulse.
Our superconductors are almost good enough, but our high-power lasers are way too inefficient and bulky. We can't even make economically viable ICF on Earth with current lasers.
The biggest un-realism in The Expanse is the lack of huge heat-sinks, at least in the show. (They aren't mentioned in the books but I assumed they'd be there.) Without heat-sinks even if the drives were >95% efficient the ship would melt. Also in the show the thrust plumes from the engines are portrayed as looking like grill flames. In reality they'd look more like beams of light fading off into space.
https://en.wikipedia.org/wiki/Heavy_ion_fusion
the argument is that "highly efficient laser" might be an oxymoron and if you were serious about commercial fusion you might trade lasers for much more efficient particle accelerators that run at a viable shot rate. Trouble is that you need heavy ions (lead) at 8GeV and it takes multiple barrels that are a km long or so... A huge machine that might be competitive with lasers for a commercial power plant but that can't be built in a subscale prototype. It might not be compatible with the magnetic nozzle though as 8GeV is not relativistic for lead ions.
A heavy ion power plant is possible in terms of the physics but needs to be the scale of a fission power plant to work at all and is projected to cost maybe 2x what an AP1000 costs assuming everything goes well and we know things usually work worse than you expect. So nobody is interested in funding a full-scale prototype, a reasonable development plan is you build several linac barrels and a test fusion facility and expect to rebuild that and add more barrels. It probably costs about what Musk spent on Twitter in the end.
K. Sandner and H. Ritsch, Physical Review Letters (2012); Temperature Gradient Driven Lasing and Stimulated Cooling
Hypothetically you could go faster if you used fission or fusion energy but practically the chemical bonds get in the way. Even
https://en.wikipedia.org/wiki/Nuclear_thermal_rocket
is limited by the strength of chemical bonds of the reactor so it gets a factor of 2 or 3 or so on exhaust velocity compared to a chemical rocket.