Interesting question. I'm not certain that you should be monitoring the health of the network, most lore about spy/terror orgs says that the more independent a cell is, the better.
A bottom-line spy knows, ideally, only his immediate superior, and the superior in turn knows only the spies he inducted and his own superior, creating a tree. That way, damage is minimized if one spy is captured.
This, on the other hand, lets you abstract the bottom-line cells, and have the lieutenants of the character be the only NPCs you have to maintain.
At this point, you can also simulate the underlings passing correct info, bad info (maybe they tortured someone to get data and the victim lied to end the suffering?), without having to actually roleplay and remember the quirks of 30 people. I guess these could be rolls done by the DM to ascertain how much info the spy network gets, or whether some people are discovered and what happens to them...
So what you have to maintain is:
- state of the immediate underlings
- state of the underlings you want to modify in your current session
- information you want to disclose to the spy network.
EDIT Nowadays this structure is used, as I said, for the security reasons I mentioned. On the other hand, in a pre-telecommunications age, this setup also means that you, as a DM, can concentrate on maintaining only local state. It's possible to have messengers that update the spy with news from a distant place, certainly, but what you don't have is constant updates on state of the whole net. Intelligence takes time to be gathered when you have only horses and -sometimes- rituals/scrying spheres for telecommunications.
After a while, the spy will have a more-or-less valid map of the net, based on both the information disclosed and the time passed from each disclosure.