To clear up what he's saying, and I understand it. Lets describe how the tracks and cylinders are set up. Unlike a CD rom, which has a single spiral 'stream' of data which starts at the inner part of the writable surface and moves outward, the HDD goes in reverse, the outer track is cylinder zero, and increments upward as it moves inward. Because the sectors have a fixed density (bits per unit of measurement), and because the outer edge moves faster than the inner edges, the sectors are all of a fixed length, thus the outer cylinders can hold more sectors than inner cylinders. This means that for one rotation, you have a much larger amount of data area running under the head.
That means cylinder 0 can read data faster than... say, cylinder 500.
Now, how does this affect performance? Well, lets say we have a 1TB HDD formatted into a single 1TB C: partition on one side, and another computer with a 1TB HDD formatted into a 120GB C: partition and the rest into a single D: partition.
We all know that Windows tends to throw files willy-nilly at the drive, files can be anywhere. Lets say on the single partitioned system that windows decided to toss the paging file somewhere stupid, like the last quarter of the C: drive. While on the other (dual partitioned) side, windows ALSO decided to throw the paging file on the last quarter of drive.
The dual partitioned drive would be MUCH faster when accessing the paging file, for a variety of reasons:
1) Because the dual partitioned C: drive is only uses the first 10% of the drive, it is going to be on the outermost rings, which move much faster under the head.
2) Because the outer rings also hold progressively more data in fewer tracks, the head wouldn't have to move back and forth as much from cylinder to cylinder to access the data.
Conversely, the single partition drive would be penalized for both those reasons. To access the files during something like a windows boot up, it would have to span the entire width of the drive to access the necessary files (those little msecs add up) and access slower sections of the drive to boot.
So - as mentioned, it won't affect reliability or lifespan, but performance can be severely affected. I found a graph where a guy ran a traced sequential disk write across his 500gb drive. The sequential write speed went from 90MB/sec down to 50MB/sec. That's about a 40% drop in performance from one side to another. Obviously a faster drive would have a higher number, but the physics still pretty much bear that all out.
And of course, a SSD has no such issues.