Thanks to those who commented on the Botnets post.
I should have pointed out a few things- as I was pointed to the Folding@home project as a "been there, done that" example.
Its certainly true that co-operative distributed systems over IP are nothing new. The net is full of them. Folding@home is just one example. They are a worthy endeavor and make large parallel processing systems available to many laudable projects that could not otherwise access such systems. They are also highly inefficient and cumbersome, relying, as they do, on the unpredictable nature of the individual processors within the group.
The reasons for this are many, so, rather than go into lengthy detail, I'm going to assume some knowledge of distributed systems and just hit the high points:
The group is in a constant state of flux, both as to processing capability and numbers. This necessitates a command and control overhead that is disproportionate to the actual computing traffic. As members are unpredictably going off and coming online, task redundancy is burdensome and further reduces efficiency. Finally, unlike a proper distributed system, there is no OS. Instead there are hundreds of small programs trying to mimic one.
What I had proposed was quite different.
While the "good guys" have to get permission and be careful not to over utilize the processing that is gifted over to them, the "bad guy's" are under no such constraints.
By utilizing a kernel level distributed system and presenting the "user" with whatever is leftover, you increase efficiency and decrease the necessity for redundancy by several orders of magnitude.
People are used to shutting down their boxes and seeing "please wait while windoze installs its updates" or some such thing. All sorts of nastiness can be done at the Kernel level-including installing a completely foreign OS- under cover of this convenient splash screen.(and that's only one of many possibilities...) The point was trying to make is that it could be done, not a detailed account of how, as this is not a technical white paper, it is after all just a blog.

I kept this post serious by design, as I did not want to infer any disrespect to those who commented publicly or privately. I will return to my usual form next time.