Actually, Bussard's polywell fusion would work great
in space, considering that the enormous vacuum chamber a test reactor would require would be unnecessary, and so long as your radiators can dissipate all the waste heat, you could build some kind of space-cruiser ship without much in the way of launches by launching the magnetic coils in an IKEA-style "flat-pack" configuration, and unpacking them on orbit.
It's still "hot" fusion, but the only technical hurdle to this is scaling it up to critical volume is expensive,
if you need to do it where there's air. Small scale test reactors work, and work reliably, and are even used as neutron sources sometimes - but power is proportional to the square of volume or something, and there's a critical volume where any smaller reactor cannot possibly reach break-even.
If the Rossi E-Cat works* - whether it generates energy is important, not the physical process through which it does - you could build a very compact, fast spacecraft with Roger Shawyer's superconducting, second-generation thruster which is supposed to generate (according to IBTimes)
"produce thrust many orders of magnitude greater than that observed by Eagleworks" with an implication that it operates at the same power level. In that case, we now have a rather compelling power-to-thrust ratio, and our choice of electric power sources with Russian sodium-cooled satellite fission-reactors as a backstop if none of the lower-mass fusion reactors are ready in time.
*(I'm not going to rule it out - but I'm not going to hold my breath, either.)