A few things to keep in mind:
1: Processor speed is not comparable across different CPU architectures. At a simplified level, "Clock Speed" measures the tick rate of the processor, and IPC (Instructions Per Clock) measures how much work a CPU does per tick. Multiplying the two gets you an approximation of how much work a single core can execute. Problem being, IPC can't be directly measured, only estimated (EG: For many generations, Intel was about 20% faster then AMD at the same clock).
2: As for Cores/Threads, a Core can run multiple threads in some CPU architectures via a technique named Simultaneous Multi-Threading (SMT). In theory, this can double performance. In practice, some CPU resources are shared between the "physical" and "logical" CPU core which limit performance gains. On Intel CPUs, they use a very small form of SMT that they call Hyperthreading which gives ~20% the performance of an additional CPU core. As a result, an 8/8 CPU would be expected to be faster then a 4/8 CPU if all other things are equal.
3: Not all software can take advantage of additional CPU cores. As a result, you will not see linear performance gains as you add additional CPU cores. For most tasks, adding more CPU cores helps by allowing more applications to run at the same time without reducing performance, rather then making a single task execute faster.