If I remember correctly rosetta@home is different to WCG. What we are doing here is using a specific rosetta version, 4.22 from what you say to fold a load of proteins.
The rosetta@home project is trying to do is to improve the quality of the results the rosetta application gives, basically evolving the application to make it either more accurate or more efficient.
The developments made in the rosetta@home project will allow for better versions of the rosetta application to be used for future projects including most likely phase 2 of the HPF project when it starts here.

So really both are worthwile, however if you have limited power to spare it is probably best not to spread yourself too thin over numerous projects.

In addition to lawrencehardin's explanations, you can also have a look at

http://www.hyper.net/dc-howto.html

which is has a special focus on life-science projects.

Personally, it's still not clear to me how exactly Folding@home's research is relevant any diseases beyond misfolding / aggregation diseases like Alzheimer's. To me, it seems a bit of a stretch for F@h to claim that simulating the fold of a protein equals doing cancer research. Especially since nowadays people around the Internet use the catchword "cancer" (which understandably gets everyones attention) to recruit new members.

Although I do understand that it's often the only way to get people involved, vs talking about "basic medical research" ...

But wouldn't HPF2 studying cancer biomarkers actually be MORE relevant to understanding this complex disease?