Testing i7-3770K - RAM variations & OCing

[CGKIII]

So, I was just lucky enough to pick up an i7-3770K for $280, and wound up building a new box. Hoping for some confirmation on a testing approach to optimize.

My primary work is likely going to be focused on the Conjectures 'R Us side of things, running PFGW tests off a PRPnet server I've got set up on another box.

Running Ubuntu 12.04, MSI-77A G45, currently have two sticks of 8GB DDR3 Corsair Vengeance 1600 MHz memory available (http://www.tigerdirect.com/applicati...FQVgMgodyS8Abw)

Considering adding another 2 sticks, but not sure if it will make a difference. Additionally, I'm interested in exploring OCing, since I've never done any of that before. I've got a 1 TB 7200 RPM 64MB 6Gb/s hard drive, unsure if there will be any performance boost by adding in a SSD (never worked with those at all before).

The idea is to set up the PRPnet server with the same 800 candidates, and look at average/total runtimes for the following scenarios:

1-4) Stock Speed:

1&2) Hyperthreading enabled, 8 clients running 100 tests each

1) Single 8 GB stick of Corsair Vengeance 1600 Mhz
2) 2x8 GB = 16 GB

3&4) Hyperthreading disabled, 4 clients running 200 tests each

3) 8 GB
4) 16 GB

5-8) Overclock the processor, same variations as above. Probably a few rounds of these at various levels of OCing.

9-12) Stock processor, overclock the RAM, same variations

13-16) Overclock both the processor and the RAM

My guesses are that performance will be better without hyperthreading, and that the 8/16 memory difference will matter, but I'm unsure of the interactions with the OCing. If the memory difference is significant, I'll consider getting another set (though maybe 16 GB is the "magic number" and there are significant diminishing returns after that).

Does this sound like a reasonable way to go about things?
Am I missing anything important?
Will an SSD result in a performance boost (likely given an OC'd proc)?

Thanks a lot.