It's actually a good question because the average user won't really ever see where the speed and additional memory capacity gets used.
The faster the computer is, and the more memory it has at hand, the more technology can be abstracted away from the users and developers.
Ease of use is the Operating System's job. If you boot up a DOS with Windows 3.1 disk, you'll find it runs pretty fast. BUT it is also lacking a lot of the comfort features that make computing easier. As speed and memory constraints get dealt with in many ways (price, new tech, etc) more of those ease-of-use functions start to appear.
To a lot of regular "desktop" users, this may not be very visible. But for example, look at graphic apps. Try to duplicate what you can do on a modern graphic app on a Windows 3.1 computer. How about something as simple to use as Blender? If you can do it, it'll take you a very long time, because without the memory and speed, you can't abstract enough of the hard work away.
The only thing that trumps clock speed are features ("tricks" in some cases perhaps) like hyperthreading, super computing, multiple cores, etc.
We can't speed our brains up. We can make a gentler learning curve, and make all the functionality of the software faster, smoother, easier. And that requires memory, speed and multiple threads.
It's worth noting that early QNX and Amiga operating systems actually did extraordinarily well with very few resources. However, maintaining them as the hardware improved was very intensive and expensive.