You didn't say whether these are idle or load temperatures. I assume they are load. If that is the case, and your room-ambient is 84F, those temperatures aren't so terrible -- depending on your over-clock setting. If these are stock-speed and and voltage settings, then the temperatures should be lower -- even at 84F.
I want to be kind. But why are you spending money on these two heatsink-fan assemblies, when Anandtech had published several reviews through 2007 and later with a cumulative comparison of products that included quite a few heatpipe coolers?
Even the G0 stepping has a thermal wattage of 95W, and if over-clocked, the actual thermal dissipation would be higher than that. But the thermal resistance of coolers like the Ultima 90, some of the later Zalman heatpipe coolers, and several designs (like one from Noctua) that lay flat to the motherboard (if that was the reason you chose the ThermalTake) -- have thermal resistances low enough to keep your temperatures down.
By comparison, with either an Ultima 90 or TRUE cooler, I had my Q6600 B3 stepping (not as efficient as your G0) showing load "PRIME95" core temperatures averaging around 66C at near-80F room-ambient running a VCORE voltage as high as 1.42V and clocked to 3.2 Ghz.
But the "CPU" temperature you cite comes from the legacy TCase sensor, which is the basis for the processor's thermal spec. It is well within that spec -- if indeed you are citing your temperatures under load conditions. For the Q6600, it can lag behind the "core" temperatures by as much as 15C degrees, or something on the order of 10C (comparing to an average of the cores).