Is your question simply academic interest, or is it to determine some action on your part?
Today, modern processors are about comparable in how much work they do per cycle.
In the past, that was not so. Advances in architecture improved the performance per cycle.
With the next 11th gen intel seems likely to advance the efficiency by perhaps 18%.
Good news there.
If a app is designed to only do one thing at a time, then the performance of a single core/thread is all important.
In the example above by USAFRet, the 65mph Honda is going to do the job faster if you only need to transport 1 to 4 riders. It is a relevant example if you are a gamer since few games can make effective use of mote than 4-6 threads.
OTOH, if your work is such that it can be divided into many threads such as editing and rendering apps, then the school bus example is relevant.
We now have from ryzen many cheap thread processors.
There also becomes a limit due to the Amdahl law which shows how the performance of the single master thread
impacts the ability to get the job done.
I think this is an under appreciated aspect of performance.
https://en.wikipedia.org/wiki/File:AmdahlsLaw.svg
Higher clocks are accompanied by higher voltage and temperature.
It make sense to not clock higher unless there is a need.
It is common for a cpu to loaf at 20% speed until there is a need for more.
That is when the faster turbo algorithm kicks in.