That's because that's pretty much the basis of 64-bit operation.
The main reason for 64-bit to come into existence, was to break the 2^32 barrier (4GB). 2^64 is 16 exabytes (giga -> tera -> peta -> exa, jsyk) and provides a lot more headroom.
I understand your cynicism in your post, regarding memory. It is over emphasised as a value of 64 bit, and I would guess that the majority of users in the world rarely need more than 2gbs!
However, there is quite a lot more invloved than memory. The first, maybe nothing to do with performance at this stage, is that we are definitely moving into 64bit technology, as we moved from 16bit. It is the way forward. New software and hardware, in the very near future, will only support 64bit.
We have 32 digit streams and 64 digit streams. Twice as much data can use the stream, in 64 bit or, to put it another way, the same data can process at twice the speed.
But it has to be kept in mind that, in computers,to take full advantage of the 64 bit stream, you need the proper gear. (Processor, software, even redesigned motherboards. Here I have to contradict my first remark, to some extent. A 64bit CPU can address more memory. If, as I said, most users do not even use 2Gbs, how would this help? Well, I am referring maybe, only to the Microsoft version. but, with this ability, the computer uses the hard disk less for storing immediate operations (swapfile or whatever), and this results in a very significant increase in speed. At the same time, of course, more memory has a big impact.
Basically, it is very impractical to begin partial updates for a proposed move to 64bit. The best plan for the average user is to wait until their old equipment is "dying" on them, and then go for the full package. 8gbs of memory, for example, would really get outstanding results on a full 64 bit computer.