i think the
hope would be that you get a thermal mass (see other reply) that is effectively variable.
with low amounts of incident energy (or when just starting) a large amount of water will heat slowly and you will get no significant steam pressure until the entire mass of liquid is boiling. but with this technology, you will get some steam almost immediately, from the water around the particles, even though the vast majority of the water remains cool. so you have a system with less lag, which is going to improve control / response.
what is not clear from the article, though, is whether this will still work well when there is a large flux of energy. hopefully both cases (traditional and particles) work equally well. in which case particles are a net win. but it may be that particles introduce new problems with "high input".
[the hope is that you get this flexibility "for free" (no need for "tubes") and yes, it works because the particles are effectively black. i have no idea how it would help remove the need for heat exchangers, though (unless they are externally powered - heat exchangers pre-warm water, which is not necessary with particles, but still makes sense if you are using waste heat, as it reduces losses and so increases efficiency).
the press release talks about making solar power more useful for small scale users in developing countries. my guess is that this helps there because you can get better efficiency when not operating at peak power, so "amateur" solar power is more efficient. the argument would be that at low power you get some stream with particles. without particles, in theory, you should also get some steam if you wait long enough (for things to heat up), but in practice, thermal losses might mean that you never do heat up enough for things to work (the heated water cooling as quickly as it is warmed).]