first one, botnetting, is proabably already thought of once,
it's simple, more like a question, a prog that 'steals' cpu and memory, (the costs to run the process)
and some bandwidth to communciate with the host,
as is said proabably suggested already, but why was it (probably) rejected?
Actually though this might have been suggested before I don't recall it, at least not in the way I interpret it - let's give it a bit more flesh...
Fact: Programs, particularly v100 + programs suck a significant amount of resources.
Fact: We have a lot of servers out there which can be used fundamentally only in two ways - 1) Install Virii to collect cash. 2) Used as bounces to hide our path through the net/or sniffer nets to find out who has been going through the net (information hubs)
Fact: A young and eager player who gets his hand on high level software is unlikely to have the hardware to run it.
Fact: Today distributed computing is possible - think of things like the SETI project where thousands of computers run part of the program.
So my interpretation on this is to have a means, probably through something like a botnet virus to allow the use of the hardware resources which are laying around unused.
Things to consider:
1) Distributed processes should cost a little more than running local - e.g. 10-20% more or to put it another way, a server with 1000 cpu (lets say there is lots of extra memory) would while running at full capacity only allow 900-800 effective CPU to be contributed to the botnet project.
2) You always need bandwidth for this. Always. One nice way to solve this is to have the botnet virus itself use a certain amount of bandwidth, and be able to manage only a certain amount of "slave" resources (memory and cpu) e.g.
Botnet virus
v 0.1 stats:
2 bps
100 cpu
5000 memory
100 MB
Can effectively "slave":
1000 cpu
50000 memory
all factors scale up linearly with the version of the botnet. (Yeah I know that's not how bots REALLY work - but it would fit into the game environment/balance that way)