It's prevented two exploits on a device I own.
One just neutered the exploit (it just didn't work) and the other caused the kernel to panic and crash. So yes, it's stopped two for two that I've seen (this is in the last ~6 years)
That's the thing though, can you prove that you haven't been exploited several times that you didn't notice and grsec which grsec didn't mitigate?