I've asked on the project forum under science for more details.
The project is very new, like any other project, and they are adding information slowly, developing the application was more important than maintaining the web page.
More information is available under the scientists home page at
http://www.iac.rm.cnr.it/~gianni/index.html
I am a participant and running this appicataion on my PS3 from the project,
A model of membrane channels: Gramicidin A
Gramicidin A channel solvated in a DMPC lipid membrane (not shown).
The permeation of ions in protein channels is important to control cell activity. However, a direct measurement of the conductivity via molecular simulations is not possible because the time scales involved are too long (micro-milli seconds). Rather, the potential of mean force (PMF) of the crossing is computed and the conductance estimated from it. We are currently looking at a simple non-equilibrium method to compute the PMF which is based on the Crooks non-equilibrium relation [Phys. Rev. E 61 2361 (2000)] over the Gramicidin A membrane protein. The equilibrium PMF is reconstructed from a set of controlled non-equilibrium pullings of ions trough the channel. This procedure also allows to compute the position dependent diffusion coefficient.
To me, from what I could read, a lot of information is in French (which I can't read), this is real science.
The project only has let a few volunteers in initially, as any project does, but they have more waiting to join. There will not be as many participants as other projects as the application only runs on the PS3 Cell Processors with Linux.
BOINC has been attempting to get Sony to add native support for BOINC (not requiring linux as Sony did for folding@home. PS3GIRD will be adding to this attempt.
---
BOINC Berkeley will be adding support for this platform, and I imagine some other projects will begin to develop applications that can take advantage of this. Since it is a new platform it just takes time for full support to develop.