There was a risk, though I gather not a big one, that one or more of the patents that hadn't been rejected yet would have been upheld. I don't know a whole lot about patent law, but there may also be an issue because the patents were in force during several years that RIM was using the relevant technology, and perhaps the rescinding of the patents wouldn't clearly release RIM from an obligation to pay for the use during the period they were in force. Probably the biggest issue is that the prospect of never-ending litigation and uncertainty -- NTP was certainly going to appeal the rescinding of its patents -- was putting such a big financial drag on RIM, and had recently started causing customers to sign on with competitors, after which inertia and upfront costs of switching would undoubtedly cause a lot of them to stay there. It probably ending up looking like it would be cheaper in the long run to pay now, even if they had eventually won.
I do think the whole affair highlights an urgent need to update both patent law and the process of patent examination.
The light goes on... that makes sense.
What RIM needed most right now was some certainty. They were starting to suffer in the marketplace merely because of the uncertainty about their service continuity.
Even if all of NTP's applications are ultimately thrown out (which appears to be on the way) RIM was better off to just settle *now*, to stop the hemorraging.
They could conceivably go back to court later with an action against NTP if the bottom falls out from under their patent apps. I wonder if that's possible. In a just world NTP should pay a price if their harrassment was all just a nuisance suit from the beginning.