For comparison of where ARM CPUs are up to, today I had the opportunity to run the same very CPU intensive program on two Raspberry Pi computers and on one of my x86 computers. As far as I know, the program is not able to use more than one core i.e. single-threaded only. Here are the results.
Pi 2B Rev 1.1 - Cortex-A7 900 MHz - 2011 - runtime: 8587 seconds
Pi 4B Rev 1.2 - Cortex-A72 1.5 GHz - 2016 - runtime: 959 seconds
v.
x86 laptop - Intel i7-7500U 3.5 GHz - 2016 - runtime: 110 seconds
So in the 5 years between 2011 and 2016, ARM has achieved a 9x speed up, while Intel by comparison flatlined, but Intel was ahead by about 9x. (For those who are into this kind of thing, the x86 hit 68 degrees C during the run, while the Pi 4 stayed a more moderate 43 degrees C.)
There are numerous objections to the above comparison - and I am probably already aware of most of them. It is presented just as a point of information. Certainly raw CPU speed is not the whole story as far as getting a responsive computer to use. The Pi 4 seems reasonable for personal use. (I’m just setting it up for someone else.)
As a digression, for this particular program the results are needed on the Pi itself but it is, in this case, possible to run the program on a much faster x86 and then copy the results across to the Pi, which I would normally do. I would not normally wait for over 2 hours for a Pi 2B to do the calculations itself !