In another thread you may have seen my comments on a similar point - from your post I decided to rebuild using i386 (for the first time) to see if that made any difference. I have had a crash a couple of times in AMD64 and now once in i386, overnight, which I never used to get prior to RC2, but the core hasn't been running long enough to see for sure that the stuttering comes in after time - I certainly was getting exactly your issue under AMD64, and I have never had that before. NB: i have bitched and whinged about stuttering video for a long time under every build so far in many threads, but that is a static amount of stutter that never got worse and is clearly unrelated to this issue (i think it is a hardware issue and I am going to rebuild with a new video card later this week).
I haven't tried the reload instead of reboot yet - if it starts happening under i386 again i will try that and get back to you, I have a gut feel that it is still happening under i386 and will get bad enough to be sure if I leave it running long enough.
I ran the grep and got a count of 165 under i386 (never tried it on AMD64), did a reboot and ran it again and got 162 after a couple of minutes.
I definitely get the impression that RC2 is implicated here, but i can't put my finger on exactly what, but i do think we need to keep this thread going, as you say. Something is going on.
BTW gazlang, i had already tried your suggestion of useevents prior to RC2 to try and solve my other stuttering issue - it didn't fix it, but i left the option in there anyway, and it definitely did not cause this increasing stutter effect, I only noticed that after going to RC2
As an aside - at least once when the core "crashed" under i386, the interface was completely frozen and I couldn't browse the Internet from the internal network because DNS on the core wasn't working, but interestingly, routing to the external network and internet was still working fine!