Freeland recognized early on that shielding all of this radiation would result in an unworkably massive vessel, so he took the opposite approach, designing Firefly to dump as much of the damaging radiation as possible directly into space. The vessel is long and pointy to minimize its radiation exposure, and most of the tail end of the ship is basically a giant radiator to handle the waste heat that can't otherwise be avoided.
The fuel tanks follow the principle of the Saturn 5 rocket, shown beside the ship for scale in the graphic above). Why add extra structure if the tanks themselves can be the structure? Michel Lamontagne's tank design does away with the structural spine common to most other designs and reduces structure to the barest minimum. Like Saturn 5, the forward part of the ship is mostly tank. The payload resides in the middle area, just before a large radiation shield that protects it from the drive's radiation.
Firefly is a big ship, although small by interstellar ship design standards. Like all the current Project Icarus designs, it would accelerate out into space, coast for a number of years, then decelerate at the Alpha Centauri system and release a swarm of sub-probes, all within a travel time of 100 years. This is faster than needed, really; increasing the travel time to several hundred years would reduce the technical challenges tremendously.
But from a practical standpoint, a 100-year mission lies at the very limit of organizational support. The Daedalus team, for instance, designed their 1978 probe for a 50-year flyby mission, because this was deemed the upper limit for a NASA engineer's professional career. Practicality in design extends not just to the technology, but to the social support necessary to make the mission possible.
There are still some important unresolved issues with the Firefly design:
The viability of Z-pinch fusion has yet to be confirmed, though Sandia National Labs and now NASA's Marshall Space Flight Center are undertaking lab tests.
The electrodes needed to supply power to the pinch must be constructed of advanced composite materials that have seen very little laboratory testing.
The direct energy conversion system needed to recapture energy from the exhaust stream to sustain the pinch is purely theoretical, with few published papers.
The high-temperature beryllium phase-change radiators that Michel Lamontagne designed for Firefly haven't yet been studied due to the expense and toxicity of beryllium. Similar systems using lithium have been used in fission reactors, but 2000K beryllium necessitates advanced piping materials (like zirconium carbide) to resist corrosion. Such systems need to be studied in a lab.
The system for communications between the vessel and Earth is rather vaguely specified at present.
Modeling the real behavior of a fusion drive exhaust in magnetic nozzles is another area where research lags considerably behind the needs of starship designers. Real world experiments -- necessarily in the large vacuum of space -- are needed to obtain data. The use of VASIMR on the ISS -- as a booster to keep the station at the proper orbit -- is a step in the right direction. (VASIMR will include the first magnetic nozzle "flown", or actually used in space.)
Design is an iterative process, and next year's Icarus Firefly may have changed considerably from this year's. Some of the prettiness may disappear if the numbers require it. But if Robert Freeland gets his way, two things are certain: it will be faster, and it will be easier to build.