Another thing to note about the CPU upgrade is the all-core sustained turbo clock speeds. The 1240v3 can sustain 3.6 GHz on all four cores at once; the 9900K can sustain a whopping 4.7 GHz on all *eight*. So we're gaining a full-tilt increase of about 30% in raw speed, plus five years of microarchitectural improvements (instructions per clock, pipeline changes, branch prediction improvements, more efficient macro-op translations, reduced instruction latency on some instructions).
Older code using old instructions like Freelancer doesn't get any sort of attention when Intel and AMD do new microarchitectures, but general improvements in each uarch generation do help quite a bit. An old-style floating point division instruction (FDIV) on our current CPU takes 18-24 cycles; on the new one, it takes 14-16. On a Pentium III (Freelancer's minimum spec), it took a fixed 38 cycles.
