are always nice. My major problem with this project is that they haven't released their algorithm yet. Protein folding is not as simple as running a cryptographic cracking scheme. Each calculation is based upon the values of the prior calculation. Without a good management scheme it is not very efficient to have too many nodes of a distributed network working on one protein.
Folding in an ab initio approach is not the best approach due to the massive amounts of permutations required. This is the approach of most distributed cracking algorithms.
This leaves us with an algorithm that is somewhat smarter. Kinda like a back-tracking algorithm. However, each step requires a lot of computations, and each next step depends on its parent. This might be a good candidate for a distibuted project if there were unlimited folding parameters, but unfortunately, there is a relatively small number of possible conformations that an amino acid can take.
This might end up working out well, but it should be looked at with skepicism until they publish their algorithm. And even then, it should also be noted that even the most advanced algorithms pale in comparison to x-ray crystallography in terms of accuracy and time. True, not every protein lends itself to the cryo process necessary for a good x-ray scatter, but it is still more accurate and will be used to verify any results that the Folding @ home project comes up with.