Those Pentium D processors are the same NetBurst microarchitecture that kept AMD in business for years. The 5160 is core microarchitecture and has 2X the max floating point throughput. I have a benchmark with a table of results that has a 3.4 GHz dual-core Pentium D at the very bottom. Of course it's written in assembly language to try to max out floating point throughput on a Core 2 Duo or Phenom so it may not be fair to a Pentium D, but why treat a crappy processor with kid gloves?
My Pentium D 3.4 with HyperThread can match the performance of Core 2 Duo only when running 4 threads effectively, running a linux with good 4 thread scheduling (which didn't exist when Pentium D was still marketed). There are plenty of jobs where it runs faster with the HT disabled, but then it can't come close to Core 2 Duo, even without any specific effort to emphasize the faster aspects of the latter. The Pentium D has been around long enough that I used to run Windows 2000 on it, but one of the reasons Win2K died is its unfriendliness to multi-threading.
So your Pentium D can match the performance of a Core 2 Duo? I'd like to see that. But I don't suppose that you can test with the benchmark I referenced above because the HT-capable Pentium D's weren't 64-bit capable. The original author of the benchmark has upgraded his 32-bit code to a new 32-bit benchmark and also has a table of results on his website. How does your Pentium D hold up in this 16-threaded benchmark?