I assume you're talking about the NTFS cluster size. Sector sizes are determined by the disk firmware and can't be changed.
The advantage of using smaller cluster sizes is that there is less wasted space. A cluster size of 4KByte wastes 2K bytes per file, on average, while a 64KByte cluster size wastes 32K bytes per file. This is because statistically speaking the last cluster in the file will only be half full on average, assuming an evenly distributed range of file sizes.
The advantage of using larger cluster sizes is that there is less file system overhead for tracking clusters and files will tend to be more contiguous since there are fewer clusters in each one.
If your average file size is fairly large (say 1MB or more) then a cluster size larger than the default 4KBytes is probably reasonable. But in practice I think it's unlikely you'll notice any significant differences.
The reason I ask, is that the two iomega drives (couldn't find a review on any 4 TB externals, and these were middle of the road expensive) are replacing one still working WD 2TB drive, and one dead WD 2TB drive. I had noticed that even with the two drives sync'd for files that the one drive showed almost 1 gig less space than the other. So my presumption is that one was set to a different cluster size than the other. Unfortunately the one drive died before I could figure out why. Not that I had the drive for that short a time, but rather that both drives started out with less than 1 TB a year ago, and only recently as they both neared 1.5TB did I notice the difference.