Shadow... The id appears to be for multiproc machines. Each physical computer (box) is supposed to have it's own unique download of the fah software so the servers can keep things straight, however multiproc machines are supposed to "copy" the software to another folder, thus the apparent need to identify them by machine name.
The following is on the Standford FAQ page.
"Does Folding@home run on dual processor machines?
Yes, you can. Additional processors must run the console version (with the "-local" command-line argument if run on Windows). First, make additional directories for each processor and copy the FAH3Console executable file into each. Then configure them with the -config switch, filling in settings for each. It is very important to make sure that under the "Advanced Settings" option each copy is given a unique machine ID (from 1 to 4). The first copy will default to a machine ID of 1, so additional copies should be given IDs of 2, 3 and 4. Each may then be run out of their installed directory, using the -local switch on windows. Instances should be attached to different processors, using whatever mechanism your OS provides for this purpose (e.g. Task Manager on Windows).
We have been working on a multiple processor version with parallelized code using OpenMP. However, we have discovered that OpenMP parallelized code does not scale very well at all with processor number with our code. So we are investigating other options for parallelization, and a final date for a parallelized version is uncertain at this time."